[
  {
    "start": 0.14,
    "end": 2.64,
    "text": "Hello friends, this is the Alphalus Podcast."
  },
  {
    "start": 3.02,
    "end": 4.04,
    "text": "I am your host Tobi!"
  },
  {
    "start": 4.38,
    "end": 9.62,
    "text": "The goal of the Alphalis podcast is to empower CTOs with the info and insight they need."
  },
  {
    "start": 24.38,
    "end": 25.98,
    "text": "Plus, if you're an experienced CTO."
  },
  {
    "start": 26.38,
    "end": 32.18,
    "text": "You will laugh at the discussion happening in our Slack space where over six hundred CTOs are sharing insights."
  },
  {
    "start": 32.38,
    "end": 64.56,
    "text": "or visit one of"
  },
  {
    "start": 68.74,
    "end": 69.18,
    "text": "product"
  },
  {
    "start": 69.76,
    "end": 82.52,
    "text": "AI first and the forefront I would say one of the Rare, I wouldn't say deep tech companies if that term still exists or a native company's that then I saw recently."
  },
  {
    "start": 83.04,
    "end": 83.74,
    "text": "coming from Germany."
  },
  {
    "start": 84.44,
    "end": 89.68,
    "text": "he went through absolutely crazy ride Was acquired by click house?"
  },
  {
    "start": 89.78,
    "end": 94.54,
    "text": "Recently as many of you are some of you might have noticed And beforehand."
  },
  {
    "start": 95.18,
    "end": 103.16,
    "text": "I don't know when like Crazy zigzag growth path to where they are now Max."
  },
  {
    "start": 103.48,
    "end": 104.16,
    "text": "I'm super excited."
  },
  {
    "start": 104.3,
    "end": 104.72,
    "text": "have you here?"
  },
  {
    "start": 104.9,
    "end": 106.78,
    "text": "Yeah, thank you so much for having me Tobias very excited."
  },
  {
    "start": 107.46,
    "end": 108.64,
    "text": "Hope the introduction was appropriate."
  },
  {
    "start": 108.68,
    "end": 108.8,
    "text": "Yes"
  },
  {
    "start": 109.72,
    "end": 110.14,
    "text": "Thank You So much."
  },
  {
    "start": 110.4,
    "end": 110.6,
    "text": "okay"
  },
  {
    "start": 111.26,
    "end": 111.58,
    "text": "Very good."
  },
  {
    "start": 112.1,
    "end": 117.6,
    "text": "and so um i always start with the same question What is your notepad like?"
  },
  {
    "start": 117.84,
    "end": 119.16,
    "text": "how did she get into computing?"
  },
  {
    "start": 119.34,
    "end": 120.7,
    "text": "And uh like why?"
  },
  {
    "start": 121.36,
    "end": 129.16,
    "text": "yeah Back then, I studied business actually and realized that mobile apps on the iPhone are super exciting."
  },
  {
    "start": 129.36,
    "end": 135.08,
    "text": "And just started with friends building mobile apps really because we thought there's so many cool things to build."
  },
  {
    "start": 136.14,
    "end": 137.78,
    "text": "This is how I ended up in engineering."
  },
  {
    "start": 137.9,
    "end": 142.26,
    "text": "I did this online course by Harvard called CS-Fifty."
  },
  {
    "start": 142.3,
    "end": 143.92,
    "text": "it was like free and online."
  },
  {
    "start": 144.26,
    "end": 149.1,
    "text": "So i learned fundamentals basically on a semester broad in Singapore at The Beach"
  },
  {
    "start": 149.56,
    "end": 156.0,
    "text": "Okay, and then you decided to study at one of the best universities for IT in Germany or IT computer science."
  },
  {
    "start": 156.36,
    "end": 158.86,
    "text": "Yeah like I went to Technical University in Munich."
  },
  {
    "start": 159.14,
    "end": 164.8,
    "text": "um this was great because uh You can go there take technical courses even if you have a business background."
  },
  {
    "start": 165.3,
    "end": 169.56,
    "text": "And then also ended up in CDTM where i met One Of My Co-Founders Mark."
  },
  {
    "start": 170.12,
    "end": 172.26,
    "text": "And you still look quite young?"
  },
  {
    "start": 172.52,
    "end": 173.02,
    "text": "You're like"
  },
  {
    "start": 173.3,
    "end": 173.7,
    "text": "thirty three"
  },
  {
    "start": 174.14,
    "end": 175.26,
    "text": "beginning of thirties or something."
  },
  {
    "start": 175.68,
    "end": 185.42,
    "text": "Wow So you hit the right spot straight away, went into like a crazy business and crazy direction."
  },
  {
    "start": 185.64,
    "end": 187.08,
    "text": "And we're then acquired."
  },
  {
    "start": 187.44,
    "end": 188.72,
    "text": "that's straightforward, right?"
  },
  {
    "start": 190.28,
    "end": 191.22,
    "text": "Yeah this was never the plan."
  },
  {
    "start": 191.92,
    "end": 197.16,
    "text": "We started building length use and the traction was really good by end of last year."
  },
  {
    "start": 197.72,
    "end": 200.54,
    "text": "The goal is to sell it quickly or something."
  },
  {
    "start": 201.0,
    "end": 201.56,
    "text": "I also believe."
  },
  {
    "start": 201.86,
    "end": 210.68,
    "text": "if there are goals selling something Probably it's really hard to achieve this because you need to care and think long-term about what you do."
  },
  {
    "start": 211.94,
    "end": 217.7,
    "text": "But then the opportunity came up, which made so much sense for us that we decided to fail to click off."
  },
  {
    "start": 219.08,
    "end": 219.42,
    "text": "Interesting!"
  },
  {
    "start": 220.34,
    "end": 226.82,
    "text": "It sounds like a wet dream of every WHU student who found something, says it super quickly"
  },
  {
    "start": 227.14,
    "end": 227.24,
    "text": "etc.,"
  },
  {
    "start": 227.56,
    "end": 230.82,
    "text": "but as you said, for techies is mostly different."
  },
  {
    "start": 231.24,
    "end": 237.98,
    "text": "also for companies these days that really have product market fit, it's supposed to be different and also changed."
  },
  {
    "start": 239.12,
    "end": 247.88,
    "text": "So was ultimately a scratch on each topic or a market call where some investor told you like hey!"
  },
  {
    "start": 247.96,
    "end": 250.4,
    "text": "You've got to go agent take, build something for agents?"
  },
  {
    "start": 250.48,
    "end": 253.62,
    "text": "Or why did end up building Building NextViews?"
  },
  {
    "start": 254.38,
    "end": 257.72,
    "text": "So basically, we were originally built completely different things."
  },
  {
    "start": 257.82,
    "end": 263.44,
    "text": "Basically typical SaaS products and we managed to end up in Y Combinator with one of these products."
  },
  {
    "start": 263.52,
    "end": 272.24,
    "text": "We had traction, we had users... ...we had a fully functioning product.. ..we were traveling to San Francisco thinking that we need to scale this."
  },
  {
    "start": 272.8,
    "end": 274.42,
    "text": "How many salespeople will be higher now?"
  },
  {
    "start": 275.36,
    "end": 277.56,
    "text": "how many engineers would we hire all those things?"
  },
  {
    "start": 278.44,
    "end": 284.24,
    "text": "And then quickly realized that we weren't too excited about our customers, and also the product back then."
  },
  {
    "start": 284.74,
    "end": 288.62,
    "text": "We weren't able to charge as much as we needed it to afford a sales."
  },
  {
    "start": 288.7,
    "end": 289.98,
    "text": "let's go-to market approach."
  },
  {
    "start": 290.52,
    "end": 294.28,
    "text": "And both of these topics basically ended us shutting down."
  },
  {
    "start": 294.54,
    "end": 297.5,
    "text": "The product that we had would basically fired all our customers."
  },
  {
    "start": 297.54,
    "end": 299.92,
    "text": "This was very hard moment in San Francisco."
  },
  {
    "start": 300.56,
    "end": 304.58,
    "text": "All other teams in Y Combinator they're pushing really hard having more revenue."
  },
  {
    "start": 306.72,
    "end": 310.06,
    "text": "Yeah, of nothing basically each week and it's a rating."
  },
  {
    "start": 310.58,
    "end": 312.64,
    "text": "And then at some point opening I released chat GPT."
  },
  {
    "start": 313.44,
    "end": 320.14,
    "text": "we didn't initially jump on the topic because we thought all the PhD AI people they are much better at this."
  },
  {
    "start": 320.22,
    "end": 321.54,
    "text": "We have no idea what is going on."
  },
  {
    "start": 322.08,
    "end": 323.32,
    "text": "until we figured out."
  },
  {
    "start": 323.56,
    "end": 325.0,
    "text": "nobody has a clue What's going on?"
  },
  {
    "start": 325.7,
    "end": 329.58,
    "text": "Then um i think It was a Sunday night ten p.m.. Something like that."
  },
  {
    "start": 330.2,
    "end": 333.06,
    "text": "Clemens my one co-founder Ask mark and me."
  },
  {
    "start": 333.38,
    "end": 337.78,
    "text": "guys Like we tried so many things now If you just had time, what would you build?"
  },
  {
    "start": 338.3,
    "end": 340.4,
    "text": "And we were saying let's built AI products."
  },
  {
    "start": 340.96,
    "end": 351.94,
    "text": "This is how we started to build AI products with very simple products such as an online course generator or a first version of GitHub."
  },
  {
    "start": 352.02,
    "end": 369.52,
    "text": "co-pilots like PR bots that he could tell hey please make the font larger for this heading like very simple applications and thereby iterated, learned how it is to build these AI applications back then with much weaker models."
  },
  {
    "start": 369.6,
    "end": 374.64,
    "text": "But we figured out quickly that traditional usability tools with logging in metrics etc."
  },
  {
    "start": 375.78,
    "end": 387.32,
    "text": "won't really help engineers improve this application because they're non-deterministic and have long strings as inputs or outputs so was hard to iterate on."
  },
  {
    "start": 387.58,
    "end": 397.76,
    "text": "them are very interested in DevTools, we read the blog post of Cloudflare and Versa when they come live."
  },
  {
    "start": 398.26,
    "end": 402.14,
    "text": "We care a lot about engineers in general and love building products in general."
  },
  {
    "start": 403.3,
    "end": 407.06,
    "text": "Your product generally is kind of deterministic."
  },
  {
    "start": 408.02,
    "end": 410.54,
    "text": "not work for engine that's the wrong word."
  },
  {
    "start": 410.82,
    "end": 414.18,
    "text": "but deterministic engine to build AI agents?"
  },
  {
    "start": 414.98,
    "end": 415.78,
    "text": "For engineers."
  },
  {
    "start": 416.16,
    "end": 420.0,
    "text": "Yeah, it's basically for a space we call it an LLM."
  },
  {
    "start": 420.1,
    "end": 428.18,
    "text": "engineering platform is basically the tool four engineers who want to build agents and that they need to iterate these agents."
  },
  {
    "start": 429.5,
    "end": 435.2,
    "text": "Just as an example, you could be an engineer and you want to build a customer support agent for example."
  },
  {
    "start": 435.66,
    "end": 442.24,
    "text": "Before you go into production You probably have fifty support cases that already know from your support team."
  },
  {
    "start": 442.68,
    "end": 445.42,
    "text": "And before you get to production... ...you want to test."
  },
  {
    "start": 445.84,
    "end": 447.16,
    "text": "how does the agent perform here?"
  },
  {
    "start": 447.32,
    "end": 449.24,
    "text": "Maybe you're missing some tool coils."
  },
  {
    "start": 449.34,
    "end": 452.26,
    "text": "maybe you should adjust the prompt To make it better performed."
  },
  {
    "start": 453.26,
    "end": 458.04,
    "text": "So we built basically a tooling experimentation Tooling to test this before even hit production."
  },
  {
    "start": 458.68,
    "end": 460.8,
    "text": "Then our users, they go to production."
  },
  {
    "start": 461.6,
    "end": 466.96,
    "text": "And then it's more like a traceability product where you see what is going on in production while the user uses."
  },
  {
    "start": 467.66,
    "end": 474.98,
    "text": "and we have something that we call online evaluations Where whenever users use it We help our users evaluate What happened?"
  },
  {
    "start": 475.36,
    "end": 480.66,
    "text": "Maybe new customer support question came up That the agent was not able To answer properly."
  },
  {
    "start": 481.3,
    "end": 495.86,
    "text": "You want be alerted about this low-performing support case, and you want to again offline experiment basically the engineer on machine or laptop."
  },
  {
    "start": 496.26,
    "end": 502.1,
    "text": "Take this production example improve agent then ship it back in which is a loop that we are working for."
  },
  {
    "start": 502.22,
    "end": 507.04,
    "text": "so its traditional observability enriched with workflow tooling"
  },
  {
    "start": 512.5,
    "end": 519.76,
    "text": "and I guess cost tracking etc."
  },
  {
    "start": 520.08,
    "end": 523.4,
    "text": "also implemented model selection, implemented."
  },
  {
    "start": 523.659,
    "end": 529.68,
    "text": "And then really like an end-to-end solution to start with developing to go live."
  },
  {
    "start": 530.76,
    "end": 532.44,
    "text": "basically individual conversation."
  },
  {
    "start": 532.88,
    "end": 534.36,
    "text": "yeah awesome"
  },
  {
    "start": 534.82,
    "end": 536.48,
    "text": "yes exactly it's."
  },
  {
    "start": 537.12,
    "end": 540.22,
    "text": "we track technical data, but also lots of business data."
  },
  {
    "start": 540.56,
    "end": 544.82,
    "text": "So technical data such as latencies how often do you call a database?"
  },
  {
    "start": 545.08,
    "end": 548.64,
    "text": "These sort of things that are more infrastructure meta-data I would call it."
  },
  {
    "start": 549.04,
    "end": 554.84,
    "text": "But we also track lots of Business relevant metrics That usually find in some sort of a data warehouse."
  },
  {
    "start": 556.14,
    "end": 559.36,
    "text": "You already mentioned cost tracking group by models that your using."
  },
  {
    "start": 560.5,
    "end": 562.92,
    "text": "We have users who do their billing via length use."
  },
  {
    "start": 563.28,
    "end": 570.62,
    "text": "They want to see How many AI tokens were Consumed by their users and they may have usage-based pricing."
  },
  {
    "start": 570.7,
    "end": 577.3,
    "text": "So, They use our APIs to fetch the tokens consumed By a single user To then charge their customers."
  },
  {
    "start": 577.7,
    "end": 581.22,
    "text": "so it's a wild mix of business and infrastructure metrics."
  },
  {
    "start": 581.54,
    "end": 582.62,
    "text": "And is open source?"
  },
  {
    "start": 582.66,
    "end": 583.92,
    "text": "Yes It's Open Source."
  },
  {
    "start": 584.54,
    "end": 586.38,
    "text": "This was an early decision."
  },
  {
    "start": 586.8,
    "end": 588.66,
    "text": "We went open source from the get go."
  },
  {
    "start": 589.5,
    "end": 592.88,
    "text": "we thought if we built for engineers It should just be open source."
  },
  {
    "start": 593.0,
    "end": 598.68,
    "text": "this felt so natural For us in so many ways that we couldn't anticipate."
  },
  {
    "start": 599.16,
    "end": 604.28,
    "text": "So the branding aspect of it is, the obvious one where you thought about yeah this will help us a lot and it did!"
  },
  {
    "start": 605.22,
    "end": 609.2,
    "text": "You could just enterprise engineers working at large enterprises."
  },
  {
    "start": 609.54,
    "end": 615.12,
    "text": "they can take our products test them out run their environment without going through procurement."
  },
  {
    "start": 617.14,
    "end": 622.7,
    "text": "We have lots of adoption about the product which is for free, but at some point they will come to us and reach out."
  },
  {
    "start": 623.0,
    "end": 625.28,
    "text": "They want to use Length Use Cloud because it's way more performant"
  },
  {
    "start": 625.34,
    "end": 625.42,
    "text": "etc.,"
  },
  {
    "start": 626.36,
    "end": 631.86,
    "text": "But that also helped in a way where our users used the products each everyway."
  },
  {
    "start": 632.5,
    "end": 637.18,
    "text": "by now are so hardened we run into each and every edge case."
  },
  {
    "start": 637.34,
    "end": 641.58,
    "text": "And we get reports, bug reports on GitHub On a daily basis."
  },
  {
    "start": 641.66,
    "end": 647.06,
    "text": "there's so many of these... ...and they're directly routed to our engineering team which is then taking quick action."
  },
  {
    "start": 647.5,
    "end": 649.84,
    "text": "So both of this have been very helpful for us."
  },
  {
    "start": 650.3,
    "end": 657.06,
    "text": "I think it also makes sense if you look at the pace that the world is turning right now Especially in AI world."
  },
  {
    "start": 658.34,
    "end": 664.64,
    "text": "For native AI tool You have to somehow differentiate, right?"
  },
  {
    "start": 664.88,
    "end": 669.64,
    "text": "I mean like the big vendors are also producing software like crazy."
  },
  {
    "start": 669.88,
    "end": 672.88,
    "text": "Right and Also now agent frameworks."
  },
  {
    "start": 672.96,
    "end": 694.64,
    "text": "Like you know open AI is most likely gonna release something and traffic already released something And then observability kind of makes it a technical tool And also want to have transparency on what is happening."
  },
  {
    "start": 696.06,
    "end": 698.4,
    "text": "I find this quite a good"
  },
  {
    "start": 699.08,
    "end": 699.34,
    "text": "choice."
  },
  {
    "start": 699.7,
    "end": 707.54,
    "text": "Exactly, there's currently OpenAI and Codex but there are also Gemini as best models changing on weekly basis."
  },
  {
    "start": 707.62,
    "end": 711.42,
    "text": "by now You just wanna swap out and try the best for you."
  },
  {
    "start": 712.0,
    "end": 715.4,
    "text": "This is why being vendor agnostic is super helpful."
  },
  {
    "start": 715.82,
    "end": 722.32,
    "text": "and then circling back into acquisition clickhouses open source It's basically the same market we are selling to."
  },
  {
    "start": 723.12,
    "end": 729.82,
    "text": "This opened up this acquisition in the first place, if closed sources would probably not have happened?"
  },
  {
    "start": 730.3,
    "end": 733.44,
    "text": "Yeah I can imagine and adoption will potentially not happen."
  },
  {
    "start": 735.7,
    "end": 739.06,
    "text": "People really look critically at what they buy these days."
  },
  {
    "start": 740.24,
    "end": 748.42,
    "text": "maybe don't see an open source pool as straight away something that has to be bought But something they can try out and play with, especially for engineers."
  },
  {
    "start": 748.5,
    "end": 758.72,
    "text": "That's quite important I think And this gives a good portion of trust that people give you straight away if they see it."
  },
  {
    "start": 759.38,
    "end": 772.28,
    "text": "Basically our go-to market motion was really Branding adoption by engineers trying it on the weekend On a hobby project Just Docker compose up in the laptop Then bringing to work basically and proposing as tool."
  },
  {
    "start": 773.64,
    "end": 779.74,
    "text": "maybe in a larger organization, the manager reaches out to us and they want to procure like an enterprise deal or something like that."
  },
  {
    "start": 780.26,
    "end": 784.62,
    "text": "And thus we don't need to do a lot of outreach to land sales."
  },
  {
    "start": 785.06,
    "end": 795.26,
    "text": "basically our sales team consisted off my co-founder Clemens and Accio especially one and half people mostly processing what came in by People who are already interested into all product."
  },
  {
    "start": 795.68,
    "end": 798.52,
    "text": "so I go market in terms of sales was very cheap."
  },
  {
    "start": 799.52,
    "end": 801.9,
    "text": "this helped us grow quickly with small teams"
  },
  {
    "start": 803.14,
    "end": 806.28,
    "text": "And you then raise angel money, I guess?"
  },
  {
    "start": 806.66,
    "end": 806.82,
    "text": "Yes."
  },
  {
    "start": 807.42,
    "end": 815.36,
    "text": "So yeah we raised from three investors Y Combinator LightSpeed and back in La Sarmilla they are now GC."
  },
  {
    "start": 815.98,
    "end": 817.94,
    "text": "so we did our seed round."
  },
  {
    "start": 818.08,
    "end": 819.72,
    "text": "basically that's it."
  },
  {
    "start": 820.56,
    "end": 822.24,
    "text": "Okay cool!"
  },
  {
    "start": 822.62,
    "end": 831.96,
    "text": "You just originally planned to raise CSA December i think Selling which doesn't sound like a fire."
  },
  {
    "start": 832.04,
    "end": 833.0,
    "text": "Yes, so"
  },
  {
    "start": 833.76,
    "end": 841.6,
    "text": "it was really this moment where Mark Lemons and I we were sitting together And we were thinking our engineering team back then We were about five engineers."
  },
  {
    "start": 842.22,
    "end": 843.42,
    "text": "its way too small for what?"
  },
  {
    "start": 843.5,
    "end": 848.92,
    "text": "We are building is SDKs UI Like data platform in real-time processing."
  },
  {
    "start": 849.56,
    "end": 850.18,
    "text": "i don't know if."
  },
  {
    "start": 850.32,
    "end": 850.94,
    "text": "think of us back Then."
  },
  {
    "start": 851.0,
    "end": 853.84,
    "text": "probably forty terabytes today per production region."
  },
  {
    "start": 853.92,
    "end": 861.42,
    "text": "It's like lots off scale and we were tiny team and even though yeah agenting coding is picking up, we still need to scale the team."
  },
  {
    "start": 861.94,
    "end": 865.54,
    "text": "And for that you do need a bit more money if you want to considerably grow the"
  },
  {
    "start": 865.64,
    "end": 865.82,
    "text": "team.\"."
  },
  {
    "start": 866.36,
    "end": 869.24,
    "text": "This is why we thought okay let's do it seriously... Let's grow this team!"
  },
  {
    "start": 869.46,
    "end": 875.36,
    "text": "Let's raise some money and then we are good again on that front end where can purely focus on hiring a stellar team?"
  },
  {
    "start": 876.54,
    "end": 880.52,
    "text": "Then Clickhouse reached out and we were debating this internally."
  },
  {
    "start": 880.98,
    "end": 893.0,
    "text": "We didn't directly jump in but eventually wrote us short doc about what would happen Like, basically playing devil's advocate because it is so easy to just say like..."
  },
  {
    "start": 893.4,
    "end": 895.04,
    "text": "Did you write that or was I flawed?"
  },
  {
    "start": 895.38,
    "end": 895.42,
    "text": "No"
  },
  {
    "start": 895.52,
    "end": 895.58,
    "text": "no."
  },
  {
    "start": 896.52,
    "end": 899.7,
    "text": "It wasn't even less than one page!"
  },
  {
    "start": 900.42,
    "end": 906.68,
    "text": "Basically have a Google Doc Because the amount of the two three reasons really matter."
  },
  {
    "start": 907.52,
    "end": 915.74,
    "text": "So we basically sat together and thought what would be the benefit And its on the data platform piece and sales piece."
  },
  {
    "start": 916.42,
    "end": 921.56,
    "text": "In the end, if you look up engineering blog posts of Datadoc for example."
  },
  {
    "start": 921.98,
    "end": 929.94,
    "text": "They reinvent their internal proprietary databases on a weekly yearly cadence basically because they need to scale it further and no more features etc."
  },
  {
    "start": 930.78,
    "end": 932.64,
    "text": "And we run everything off of Clickhouse."
  },
  {
    "start": 932.8,
    "end": 934.06,
    "text": "It's basically real time."
  },
  {
    "start": 934.12,
    "end": 945.38,
    "text": "columnar database is made for observability use cases database, but we realized at some point our interest may diverge from whatever ClickHouse has."
  },
  {
    "start": 945.94,
    "end": 959.44,
    "text": "And then I am in a position where i need to hire um a database team and probably will fail it because you have to catch up with what ever ClickHouses built and build the things that we need on top or branched from ClickHaus and start developing there."
  },
  {
    "start": 959.94,
    "end": 962.6,
    "text": "It's very big undertaking and risky."
  },
  {
    "start": 963.54,
    "end": 965.36,
    "text": "The second thing is really sales."
  },
  {
    "start": 965.6,
    "end": 967.46,
    "text": "We don't know how."
  },
  {
    "start": 968.2,
    "end": 971.74,
    "text": "open source death tools sales out of germany."
  },
  {
    "start": 972.28,
    "end": 975.1,
    "text": "this is really hard and click off has done this."
  },
  {
    "start": 975.18,
    "end": 977.4,
    "text": "they have sales people that know what are composers."
  },
  {
    "start": 977.92,
    "end": 979.2,
    "text": "They can talk with."
  },
  {
    "start": 980.2,
    "end": 981.62,
    "text": "yeah it is so does."
  },
  {
    "start": 981.92,
    "end": 993.42,
    "text": "the talk was engineers from a very impressive companies, as an opportunity to basically jump on their sales team and have a direct connection with their CoreDB Team."
  },
  {
    "start": 993.84,
    "end": 1010.56,
    "text": "It's saving us so many years in developing all of this internally, which was the reason why we thought that if we go through these two points then the chance for Length Views coming out at the top is by far highest!"
  },
  {
    "start": 1010.9,
    "end": 1010.94,
    "text": "In"
  },
  {
    "start": 1016.02,
    "end": 1022.44,
    "text": "the previous question, you mentioned that there was a really hot phase before Christmas without naming names."
  },
  {
    "start": 1023.32,
    "end": 1029.0,
    "text": "What does it feel like when half of the AI ecosystem is suddenly knocking at your door?"
  },
  {
    "start": 1029.7,
    "end": 1029.859,
    "text": "Yeah!"
  },
  {
    "start": 1030.88,
    "end": 1035.64,
    "text": "It's a really busy time so don't think about how crazy this is in those moments."
  },
  {
    "start": 1035.68,
    "end": 1039.96,
    "text": "but yeah...it wasn't crazy right because we need to prepare also lots and lots of work."
  },
  {
    "start": 1040.5,
    "end": 1047.54,
    "text": "It's really a lot of work to prepare all these meetings and make the fundraise, acquisition then works."
  },
  {
    "start": 1047.579,
    "end": 1050.7,
    "text": "And this is where Mark and Clemens especially my two co-founders."
  },
  {
    "start": 1050.74,
    "end": 1052.04,
    "text": "they're really excelled at that."
  },
  {
    "start": 1052.1,
    "end": 1053.24,
    "text": "in the first very impressive."
  },
  {
    "start": 1053.66,
    "end": 1058.38,
    "text": "so yeah... And processing it happens a bit later."
  },
  {
    "start": 1059.62,
    "end": 1062.42,
    "text": "How does it feel now being part of the kick house team?"
  },
  {
    "start": 1064.54,
    "end": 1066.42,
    "text": "How's the cultural stuff like that?"
  },
  {
    "start": 1066.6,
    "end": 1068.36,
    "text": "Did you were surprised by anything?"
  },
  {
    "start": 1068.54,
    "end": 1076.52,
    "text": "or is it now just continuing to crazy riot with a very pushy team, but sales your product straight away."
  },
  {
    "start": 1076.74,
    "end": 1079.6,
    "text": "Or Is there anything worth surprising for you?"
  },
  {
    "start": 1079.76,
    "end": 1079.76,
    "text": "I"
  },
  {
    "start": 1080.18,
    "end": 1100.0,
    "text": "think in general nothing unexpected because when we met the founders of click house We realized Individual engineers make decisions and give them ownership so they can move fast."
  },
  {
    "start": 1100.54,
    "end": 1104.84,
    "text": "And just their excellent seeing how often they have IPO companies, etc."
  },
  {
    "start": 1105.14,
    "end": 1111.74,
    "text": "It's like a very impressive founder team and the remaining management team of Clickhouse is also super impressive."
  },
  {
    "start": 1111.8,
    "end": 1115.06,
    "text": "so with this nothing was surprising."
  },
  {
    "start": 1115.46,
    "end": 1124.4,
    "text": "it's a lot of fun to work with them and now we've been in Tokyo last week With entire click house organization for an off-site and I was so impressed To see what there all shipping."
  },
  {
    "start": 1124.72,
    "end": 1131.44,
    "text": "its like six hundred people building highly complex ultra-scaling infrastructure product."
  },
  {
    "start": 1131.92,
    "end": 1136.36,
    "text": "So this was very impressive to see, it feels the decision was perfect."
  },
  {
    "start": 1159.2,
    "end": 1165.62,
    "text": "Crazy productivity mode where you work like I don't know sixteen hours a day or something."
  },
  {
    "start": 1166.08,
    "end": 1168.74,
    "text": "Or how is that your view on that these days?"
  },
  {
    "start": 1168.8,
    "end": 1171.46,
    "text": "Like, i think there are two different perspectives."
  },
  {
    "start": 1172.34,
    "end": 1175.16,
    "text": "One of them makes us so much more productive."
  },
  {
    "start": 1175.38,
    "end": 1180.08,
    "text": "We do not necessarily have to work sixteen hours and the other one works sixteen hours."
  },
  {
    "start": 1182.04,
    "end": 1183.9,
    "text": "How did it change for you?"
  },
  {
    "start": 1184.48,
    "end": 1189.62,
    "text": "Honestly, it didn't change a lot for I think the engineering team basically nothing changed."
  },
  {
    "start": 1189.88,
    "end": 1198.74,
    "text": "We are still this small team we're still understaffed with still have huge scale growing about twenty percent in terms of data intake each month."
  },
  {
    "start": 1199.68,
    "end": 1204.3,
    "text": "This is still a lot to build and our users are demanding so many cool features."
  },
  {
    "start": 1205.24,
    "end": 1207.2,
    "text": "I also think it makes so much sense to build these features."
  },
  {
    "start": 1207.26,
    "end": 1211.28,
    "text": "So, its still a high velocity environment where we need to push."
  },
  {
    "start": 1211.82,
    "end": 1215.8,
    "text": "It's not that Clickhouse is now sending us an engineering team and telling you can relax now."
  },
  {
    "start": 1216.22,
    "end": 1216.26,
    "text": "No!"
  },
  {
    "start": 1216.42,
    "end": 1222.08,
    "text": "We have to hire our own teams And make our product decisions and move really fast."
  },
  {
    "start": 1223.18,
    "end": 1229.02,
    "text": "For agentic coding this took off since Christmas and models became better."
  },
  {
    "start": 1230.44,
    "end": 1240.6,
    "text": "The Velocity per engineer has increased But it's not in a state, I would say where i'm waking up in the morning and is just done."
  },
  {
    "start": 1240.64,
    "end": 1242.28,
    "text": "And then can merge major PR."
  },
  {
    "start": 1242.76,
    "end": 1245.88,
    "text": "there still so much to review an understand of what going on?"
  },
  {
    "start": 1246.9,
    "end": 1261.84,
    "text": "The next step is to understand how really review things make sure that they don't blow-up production because unperforming SQL query shipped or some design patterns are not implemented correctly, these sort of things."
  },
  {
    "start": 1261.9,
    "end": 1266.1,
    "text": "So like an engineer still needs to take the time to really understand what happens."
  },
  {
    "start": 1267.34,
    "end": 1273.76,
    "text": "so it's not in a world where basically everything is for free and we can just ship at ultra speed."
  },
  {
    "start": 1274.5,
    "end": 1276.12,
    "text": "It would potentially also be boring right?"
  },
  {
    "start": 1277.64,
    "end": 1281.34,
    "text": "Yeah I mean that's the question how does building products work?"
  },
  {
    "start": 1282.76,
    "end": 1289.14,
    "text": "Is there actually like a software development life cycle where you just write specs and get the result in the morning?"
  },
  {
    "start": 1290.44,
    "end": 1291.74,
    "text": "I'm not sure, i haven't seen that really."
  },
  {
    "start": 1292.2,
    "end": 1306.16,
    "text": "In The Wild it's an interesting idea or good idea but somehow accelerate from A to B. And i think exploration has also something."
  },
  {
    "start": 1308.06,
    "end": 1310.66,
    "text": "Yes, but it's also heavily in AI enabled."
  },
  {
    "start": 1310.76,
    "end": 1315.8,
    "text": "I mean we get so much user feedback on GitHub discussions for example its tons of threads."
  },
  {
    "start": 1316.54,
    "end": 1317.22,
    "text": "I want to build this."
  },
  {
    "start": 1317.68,
    "end": 1318.9,
    "text": "What do my users tell me?"
  },
  {
    "start": 1319.04,
    "end": 1323.4,
    "text": "And then let Claude co-work, give him a summary of what users actually"
  },
  {
    "start": 1323.46,
    "end": 1323.68,
    "text": "want.\"."
  },
  {
    "start": 1324.18,
    "end": 1326.88,
    "text": "I read an RFC from one of our engineers yesterday."
  },
  {
    "start": 1327.02,
    "end": 1327.76,
    "text": "he did exactly that."
  },
  {
    "start": 1327.86,
    "end": 1338.6,
    "text": "it was the perfect summery off you what use us wants and Then From there You can again work with Claude or codex To figure out what's The best way to implement This and then iterate also on this plan."
  },
  {
    "start": 1338.8,
    "end": 1346.6,
    "text": "so the planning stage is already highly AI enabled, and then giving this into a context to an agent who execute."
  },
  {
    "start": 1346.64,
    "end": 1349.14,
    "text": "on top of that is basically just the linear integration two."
  },
  {
    "start": 1349.76,
    "end": 1352.62,
    "text": "Sure cloud code working off these tasks."
  },
  {
    "start": 1352.96,
    "end": 1357.34,
    "text": "so I think right now it's basically The engineer who needs to decide."
  },
  {
    "start": 1357.46,
    "end": 1358.94,
    "text": "i want to prioritize This project."
  },
  {
    "start": 1359.64,
    "end": 1363.32,
    "text": "he needs To understand based On the user research what are really?"
  },
  {
    "start": 1363.38,
    "end": 1369.22,
    "text": "The requirements What's important here where at most Difficult and important decisions."
  },
  {
    "start": 1369.58,
    "end": 1374.2,
    "text": "And this is where really what tastes matters, and we're understanding the user's methods."
  },
  {
    "start": 1374.98,
    "end": 1384.7,
    "text": "for example for us We always try to build a platform which was very open and very like it could be used for so many different use cases not very specialized not verticalized."
  },
  {
    "start": 1386.0,
    "end": 1390.86,
    "text": "Having this in a mind of an engineer, let's then make the right decisions on the plan."
  },
  {
    "start": 1391.44,
    "end": 1407.5,
    "text": "And after the agent basically implemented everything it is again like review step and making sure that everything was implemented correctly ship into production whatever monitoring and production I mean, us as an infrastructure product."
  },
  {
    "start": 1407.56,
    "end": 1409.24,
    "text": "we are on call twenty four seven."
  },
  {
    "start": 1409.82,
    "end": 1425.74,
    "text": "something may blow up at three M. On a Sunday night happened often enough but then having again in agent being connected to data doc and all the tools that We have And helping me To understand what's The root cause is super helpful."
  },
  {
    "start": 1426.3,
    "end": 1428.42,
    "text": "Yeah give you A context That Actually"
  },
  {
    "start": 1428.72,
    "end": 1429.22,
    "text": "Is Already"
  },
  {
    "start": 1429.32,
    "end": 1429.74,
    "text": "Prepared."
  },
  {
    "start": 1429.82,
    "end": 1436.78,
    "text": "Yes Hey The Internet Is Down Fastly is down or Cloudflare is down and have you checked this?"
  },
  {
    "start": 1436.94,
    "end": 1438.42,
    "text": "And I already figured out this."
  },
  {
    "start": 1438.8,
    "end": 1439.68,
    "text": "This how to fix it,"
  },
  {
    "start": 1440.16,
    "end": 1440.3,
    "text": "but"
  },
  {
    "start": 1440.44,
    "end": 1443.64,
    "text": "still the human in loop basically does all"
  },
  {
    "start": 1444.16,
    "end": 1445.04,
    "text": "of that."
  },
  {
    "start": 1445.34,
    "end": 1445.78,
    "text": "right now."
  },
  {
    "start": 1446.68,
    "end": 1455.04,
    "text": "what we try do get as much context into Agent Harness give it all the MCPs All tools."
  },
  {
    "start": 1456.3,
    "end": 1459.3,
    "text": "We are very lucky here because we're open source."
  },
  {
    "start": 1461.76,
    "end": 1463.32,
    "text": "Also about us as a company, our culture."
  },
  {
    "start": 1463.38,
    "end": 1465.72,
    "text": "we have basically handbook out there."
  },
  {
    "start": 1466.12,
    "end": 1478.06,
    "text": "Our internal company Handbook is on our LengFuse.com homepage and this is super helpful because now agents can just go find the content And it's just accumulating over time."
  },
  {
    "start": 1478.12,
    "end": 1484.72,
    "text": "alot and then combining all these contents with very powerful model really helps alot."
  },
  {
    "start": 1485.14,
    "end": 1497.08,
    "text": "so I think its currently the figuring out which content you're missing to improve personal productivity, basically your team productivity and making sure all these connections are made today agent harness."
  },
  {
    "start": 1497.68,
    "end": 1499.2,
    "text": "And where would you say?"
  },
  {
    "start": 1499.28,
    "end": 1506.88,
    "text": "You stand yourself with your own cloud and observability in like this is prepared for harness-prepared context."
  },
  {
    "start": 1508.3,
    "end": 1509.98,
    "text": "Where do we understand how happy I was?"
  },
  {
    "start": 1510.06,
    "end": 1512.46,
    "text": "status quo The real"
  },
  {
    "start": 1514.08,
    "end": 1514.26,
    "text": "I mean"
  },
  {
    "start": 1515.08,
    "end": 1515.52,
    "text": "about the link."
  },
  {
    "start": 1515.58,
    "end": 1515.66,
    "text": "Yeah,"
  },
  {
    "start": 1515.88,
    "end": 1520.4,
    "text": "exactly so i mean everybody reads on linkedin and x each day."
  },
  {
    "start": 1520.66,
    "end": 1520.82,
    "text": "how?"
  },
  {
    "start": 1521.74,
    "end": 1523.44,
    "text": "Crazy productive everybody else is."
  },
  {
    "start": 1524.06,
    "end": 1530.52,
    "text": "i think we are also not happy About where currently at we could be much better And we're investing there."
  },
  {
    "start": 1530.62,
    "end": 1535.22,
    "text": "it's basically projects that are under a quarterly roadmap of our engineering teams."
  },
  {
    "start": 1536.76,
    "end": 1539.7,
    "text": "So no, I'm not happy because i think we could be doing better."
  },
  {
    "start": 1539.82,
    "end": 1543.8,
    "text": "We have lots of ideas on what we can improve and where do want to move."
  },
  {
    "start": 1543.84,
    "end": 1550.7,
    "text": "but it's also like an exciting time really understand workflows how you're more productive."
  },
  {
    "start": 1550.82,
    "end": 1559.48,
    "text": "so yeah engineers who are curious about these sort things probably had very fun times right now understanding the next step in industry."
  },
  {
    "start": 1559.64,
    "end": 1559.86,
    "text": "And"
  },
  {
    "start": 1562.8,
    "end": 1577.24,
    "text": "if someone wants to now, let's say build a similar setup to yours or like given what you learned with your own cloud building an observability stack for normal web app these days."
  },
  {
    "start": 1578.02,
    "end": 1578.88,
    "text": "What would you I mean?"
  },
  {
    "start": 1579.12,
    "end": 1581.76,
    "text": "You mentioned data don't write it or something similar and you're ready etc."
  },
  {
    "start": 1582.96,
    "end": 1584.42,
    "text": "but this is the old world."
  },
  {
    "start": 1585.12,
    "end": 1587.0,
    "text": "how do you do in the new world?"
  },
  {
    "start": 1587.42,
    "end": 1597.22,
    "text": "basically be prepared A genetic resolution, which at a certain point will also come that you automatically solve issues with an agent."
  },
  {
    "start": 1598.08,
    "end": 1598.36,
    "text": "What do say?"
  },
  {
    "start": 1598.64,
    "end": 1601.8,
    "text": "what would be your setup if we could restart from now?"
  },
  {
    "start": 1602.44,
    "end": 1607.16,
    "text": "So it's the multi-layout answer here because many things matter."
  },
  {
    "start": 1608.22,
    "end": 1621.52,
    "text": "I think The data platform under the hood needs to be blazing fast and this is where click house really excels processing billions of rows in a database sub-second, the sum off total costs for given user."
  },
  {
    "start": 1622.6,
    "end": 1629.64,
    "text": "A query like that is just responding very fast on gigantic amount data and low latency."
  },
  {
    "start": 1629.9,
    "end": 1634.36,
    "text": "database queries sound technical but they have real impact on an agent."
  },
  {
    "start": 1634.84,
    "end": 1647.06,
    "text": "if you are hooked up with LengFuse with our MCP client for example you want to have a very quick answer and don't wait for ten minutes."
  },
  {
    "start": 1655.46,
    "end": 1659.32,
    "text": "You also need the basically agentic setup,"
  },
  {
    "start": 1660.1,
    "end": 1662.58,
    "text": "but the data platform could be datadoc right?"
  },
  {
    "start": 1665.66,
    "end": 1672.0,
    "text": "So let's say I had better stack for monitoring which then emits and on-call signals."
  },
  {
    "start": 1672.06,
    "end": 1679.26,
    "text": "somewhere like my application crashes, there's an incident happening that I don't know."
  },
  {
    "start": 1679.38,
    "end": 1681.88,
    "text": "a web hook to where."
  },
  {
    "start": 1681.92,
    "end": 1682.72,
    "text": "What happens next?"
  },
  {
    "start": 1683.14,
    "end": 1684.78,
    "text": "You have Datadoc you have that..."
  },
  {
    "start": 1685.1,
    "end": 1685.24,
    "text": "Yes!"
  },
  {
    "start": 1686.08,
    "end": 1687.36,
    "text": "...and do we have that incident happening?"
  },
  {
    "start": 1687.68,
    "end": 1688.38,
    "text": "what happens next?"
  },
  {
    "start": 1689.08,
    "end": 1694.18,
    "text": "Probably you want to have it hooked up with basically the MCP of BetaStack or DataDoc, whatever."
  },
  {
    "start": 1694.24,
    "end": 1706.34,
    "text": "You'll also need to have the repositories that your code needs in context so... ...you've got the actual codebase for the servers running but ideally infrastructure codes."
  },
  {
    "start": 1706.74,
    "end": 1714.5,
    "text": "We realized having the infrastructure code in Terraform is extremely helpful because the agent knows what you have deployed and production."
  },
  {
    "start": 1715.38,
    "end": 1720.28,
    "text": "So these need to be in context, so ideally your have a webhook that could do whatever like."
  },
  {
    "start": 1720.32,
    "end": 1721.66,
    "text": "there's integrations to linear."
  },
  {
    "start": 1721.72,
    "end": 1731.24,
    "text": "I think linear it would be workflow engine for this of the future where you just create a linear ticket off our data doc alert or some of the pager tutu alerts"
  },
  {
    "start": 1731.3,
    "end": 1731.94,
    "text": "Or GitHub issues."
  },
  {
    "start": 1732.12,
    "end": 1732.58,
    "text": "take it right?"
  },
  {
    "start": 1732.62,
    "end": 1733.64,
    "text": "To keep even simpler"
  },
  {
    "start": 1734.04,
    "end": 1734.16,
    "text": "Yeah."
  },
  {
    "start": 1734.58,
    "end": 1741.84,
    "text": "And then from their Basically, you launch a Codex session or a code session to start the investigation."
  },
  {
    "start": 1742.28,
    "end": 1745.06,
    "text": "And get up issues from some repository."
  },
  {
    "start": 1745.16,
    "end": 1749.32,
    "text": "that's like an observability repository for your setup?"
  },
  {
    "start": 1749.88,
    "end": 1755.4,
    "text": "I mean for us right now it's basically we run this manually, where the alert comes in."
  },
  {
    "start": 1760.74,
    "end": 1773.02,
    "text": "I could now start, and this is the next step that we need to take as basically having a web hook and fully automate."
  },
  {
    "start": 1773.1,
    "end": 1784.12,
    "text": "So when i'm getting awake... ...I am opening up my laptop but already have the first investigation so it doesn't need to wait And then from there also starts like a fixed proposal would be our next step."
  },
  {
    "start": 1785.1,
    "end": 1798.22,
    "text": "To stay with an imaginary setup There will be dedicated AI context repository which kind of describes what your company does Describes all the mcp servers Um, and then gets like a GitHub actions."
  },
  {
    "start": 1798.82,
    "end": 1806.18,
    "text": "A web hook triggers some workflow that then collects the context likes starts codex or whatever you harness your use right?"
  },
  {
    "start": 1806.64,
    "end": 1821.5,
    "text": "So it's codex collect all the other of the contacts them And basically fires Then or wakes you up at night in The next step because it can't resolve itself Or sends you like I don't know a slack message."
  },
  {
    "start": 1821.86,
    "end": 1823.12,
    "text": "Hey, do you want me to resolve it?"
  },
  {
    "start": 1823.2,
    "end": 1823.6,
    "text": "Yes or no."
  },
  {
    "start": 1824.96,
    "end": 1825.64,
    "text": "This is what I would"
  },
  {
    "start": 1825.68,
    "end": 1825.74,
    "text": "do.\"."
  },
  {
    "start": 1826.56,
    "end": 1827.0,
    "text": "What's our plan?"
  },
  {
    "start": 1827.16,
    "end": 1827.4,
    "text": "currently?"
  },
  {
    "start": 1827.96,
    "end": 1833.78,
    "text": "run more on EC-II instances basically having a server running and loads all the contacts in there."
  },
  {
    "start": 1834.08,
    "end": 1844.62,
    "text": "Cloud code, MCP server skills everything goes in there And then you have this standing server that should receive the webhook of if something was going wrong."
  },
  {
    "start": 1846.0,
    "end": 1860.78,
    "text": "The engineer will be able connect with the EC-III instance the investigation, but then also be able to start a fix in the Infra repository or one of the app repositories that have a fixed life and production as soon as possible."
  },
  {
    "start": 1861.2,
    "end": 1869.1,
    "text": "The GitHub action is nice because it's very easy to configure But its basically stopping at some point And you need do stuff stateless."
  },
  {
    "start": 1869.56,
    "end": 1880.48,
    "text": "So having an EC-II instance which was completely isolated clots to whatever it wants in there is probably a best idea."
  },
  {
    "start": 1881.04,
    "end": 1885.66,
    "text": "And the setup we just described... Doesn't include length use at all, right?"
  },
  {
    "start": 1886.06,
    "end": 1887.28,
    "text": "Like where comes like this?"
  },
  {
    "start": 1887.36,
    "end": 1889.88,
    "text": "and why do I need to thank shoes in addition."
  },
  {
    "start": 1889.94,
    "end": 1891.4,
    "text": "Do i need it for the setup at All?"
  },
  {
    "start": 1891.66,
    "end": 1891.76,
    "text": "yeah"
  },
  {
    "start": 1892.08,
    "end": 1897.42,
    "text": "so basically we just talked about our set up of how We monitor our infrastructure."
  },
  {
    "start": 1897.68,
    "end": 1909.34,
    "text": "but if i'm now building an agent And this is also our lengthuse future Is to build an in-app agent Where we render want to render custom ui based on whatever user currently wants To have."
  },
  {
    "start": 1909.4,
    "end": 1914.92,
    "text": "So say Again, circling back to the customer support case maybe someone complained."
  },
  {
    "start": 1915.48,
    "end": 1917.68,
    "text": "Hey I just got a really bad response here."
  },
  {
    "start": 1917.76,
    "end": 1922.92,
    "text": "for this use case human agent needed to take us over and get it off of that line and save it."
  },
  {
    "start": 1922.96,
    "end": 1927.56,
    "text": "This could also be dispatching a webhook And then start to interact The same way."
  },
  {
    "start": 1927.7,
    "end": 1929.32,
    "text": "we just discussed with Datadoc."
  },
  {
    "start": 1929.82,
    "end": 1938.32,
    "text": "Interact the Same Way With Langfuse where you have all these prompts in there understand which tool calls were used by the agents, et cetera."
  },
  {
    "start": 1938.8,
    "end": 1953.22,
    "text": "And then basically do this same loop of trying to improve the agent off of negative feedback that we just got and it's basically the same use case in the same loop Just for engineers work building these agentic applications"
  },
  {
    "start": 1953.7,
    "end": 1957.9,
    "text": "meaning That you would not necessarily need length shoes for this incident agent."
  },
  {
    "start": 1958.1,
    "end": 1964.1,
    "text": "But whenever you want to go deeper or whatever you think they're I mean, for the incident agent that needs to be human in a loop right?"
  },
  {
    "start": 1964.16,
    "end": 1967.68,
    "text": "Like you basically describe workflow and code."
  },
  {
    "start": 1968.32,
    "end": 1971.14,
    "text": "And then you say at this step... ...I want to do this and this!"
  },
  {
    "start": 1971.96,
    "end": 1979.08,
    "text": "If A involves with humans if B goes straight away.. ..and this is what we would do in length."
  },
  {
    "start": 1979.2,
    "end": 1981.8,
    "text": "use like the agent would be powered or"
  },
  {
    "start": 1984.56,
    "end": 1985.66,
    "text": "not."
  },
  {
    "start": 1985.7,
    "end": 1987.22,
    "text": "Yeah it's complex world."
  },
  {
    "start": 1988.3,
    "end": 1992.6,
    "text": "Maybe for this one use case, again talking about the customer support case."
  },
  {
    "start": 1993.62,
    "end": 2002.04,
    "text": "Someone asked something completely new that we never used or optimized our agent for and maybe context is missing to the agents."
  },
  {
    "start": 2002.44,
    "end": 2005.88,
    "text": "so then action would be prepare these contexts."
  },
  {
    "start": 2006.38,
    "end": 2015.12,
    "text": "This could be like an agent preparing this or a human realizes oh I need additional MCP server or the MCP Server needs additional capabilities."
  },
  {
    "start": 2016.74,
    "end": 2023.86,
    "text": "then the engineer is looping and basically coding improvements to the agent, going into production again."
  },
  {
    "start": 2024.02,
    "end": 2027.1,
    "text": "And then monitoring with Langfuse again so that Langfus monitoring."
  },
  {
    "start": 2027.16,
    "end": 2030.08,
    "text": "it's really about how good does my agent currently working?"
  },
  {
    "start": 2030.64,
    "end": 2033.76,
    "text": "It's less about this one EC tool instance just died."
  },
  {
    "start": 2034.26,
    "end": 2041.26,
    "text": "many of our users are using Langfuce in parallel with traditional possibility To understand these infrastructure topics off."
  },
  {
    "start": 2041.36,
    "end": 2042.56,
    "text": "My database Is Just Going Down."
  },
  {
    "start": 2043.28,
    "end": 2046.82,
    "text": "We Are More Focused on The Quality Of The Agent Yeah,"
  },
  {
    "start": 2047.42,
    "end": 2056.94,
    "text": "and this then often is less the traditional observability but observability around some agent use case like hey I want to automate my customer support."
  },
  {
    "start": 2057.0,
    "end": 2059.4,
    "text": "Yes i want to Automate...I don't know."
  },
  {
    "start": 2059.44,
    "end": 2067.5,
    "text": "uh let's say in my whole customer outreach And by my sourcing engine etc..and This what you would do with lengthuse."
  },
  {
    "start": 2067.54,
    "end": 2072.86,
    "text": "while observability potentially Traditional observability Potentially not right?"
  },
  {
    "start": 2074.36,
    "end": 2074.94,
    "text": "yes!"
  },
  {
    "start": 2076.38,
    "end": 2081.58,
    "text": "I think like really pushes me to, a question that if you open LinkedIn."
  },
  {
    "start": 2082.54,
    "end": 2087.739,
    "text": "You think the world is already at in ninety nine percent agents right from your perspective?"
  },
  {
    "start": 2087.98,
    "end": 2097.84,
    "text": "Like the non-linked version where do we actually stand with agents and production today And how many companies are really succeeding there already?"
  },
  {
    "start": 2099.12,
    "end": 2101.72,
    "text": "Where does the word stand?"
  },
  {
    "start": 2102.62,
    "end": 2102.96,
    "text": "Yeah From"
  },
  {
    "start": 2103.0,
    "end": 2103.58,
    "text": "Your Perspective"
  },
  {
    "start": 2104.68,
    "end": 2113.14,
    "text": "It's obviously picking up a lot because just the models get so much better and we are also iterating on how models get context."
  },
  {
    "start": 2113.2,
    "end": 2116.52,
    "text": "So things get much better, I think in terms of products."
  },
  {
    "start": 2116.64,
    "end": 2134.56,
    "text": "there is lots of products from our customers that are super impressive, they're client-facing and work really good for example tools to find candidates for our engineering hiring with AI and they're so much better than all the traditional tools out there."
  },
  {
    "start": 2134.94,
    "end": 2140.28,
    "text": "And based on a text description or voice input, find me extremely good candidates!"
  },
  {
    "start": 2141.46,
    "end": 2142.9,
    "text": "These tools work extremely well."
  },
  {
    "start": 2143.72,
    "end": 2145.72,
    "text": "I think also have tons of traction."
  },
  {
    "start": 2145.9,
    "end": 2147.8,
    "text": "So these use cases exist."
  },
  {
    "start": 2150.18,
    "end": 2153.44,
    "text": "lots of internal operational use cases exist."
  },
  {
    "start": 2153.54,
    "end": 2158.88,
    "text": "So, lots of enterprises have manual workflows that you could just automate with agents."
  },
  {
    "start": 2159.3,
    "end": 2160.84,
    "text": "These use-cases exists a lot."
  },
  {
    "start": 2161.42,
    "end": 2163.66,
    "text": "This is where we talk about enterprises."
  },
  {
    "start": 2164.58,
    "end": 2176.6,
    "text": "And then there's also less agentic use case but very high volume uses such as scanning invoices for example and extracting data out of invoice at gigantic volumes."
  },
  {
    "start": 2177.14,
    "end": 2180.42,
    "text": "these cases exist as well and are now extremely good."
  },
  {
    "start": 2180.52,
    "end": 2191.7,
    "text": "And you don't need to fine tune a model on your use case because I have Japanese invoices or something like that mixed with English ones, no it's just working incredibly well."
  },
  {
    "start": 2193.3,
    "end": 2202.26,
    "text": "Yeah those really narrow use-cases That yeah Like You mentioned flagging invoises Or doing customer support in a SaaS company etc."
  },
  {
    "start": 2202.84,
    "end": 2207.16,
    "text": "This is i think Doing Customer Support potentially is not Yeah, I don't know."
  },
  {
    "start": 2208.18,
    "end": 2209.52,
    "text": "But it's what i feel like."
  },
  {
    "start": 2210.22,
    "end": 2212.24,
    "text": "for us customer support works extremely well."
  },
  {
    "start": 2212.3,
    "end": 2217.94,
    "text": "we have two guys doing customer support and We haven basically agents."
  },
  {
    "start": 2218.34,
    "end": 2224.0,
    "text": "that is doing first-levels customer support For Us And I think It's about Doing About Ten Thousand Conversations a Week?"
  },
  {
    "start": 2224.84,
    "end": 2231.96,
    "text": "We would never be able to do this type of Support With Two People like impossible."
  },
  {
    "start": 2232.2,
    "end": 2247.68,
    "text": "And there only a very small fraction is routed to these two guys who are technical and can help our users with deeper issues, but in this case it's really good for customers at least the feedback we get."
  },
  {
    "start": 2248.64,
    "end": 2253.08,
    "text": "so customer support already I think working extremely well."
  },
  {
    "start": 2253.64,
    "end": 2255.38,
    "text": "Okay then pens down again."
  },
  {
    "start": 2255.78,
    "end": 2256.84,
    "text": "how does that work exactly?"
  },
  {
    "start": 2256.9,
    "end": 2259.12,
    "text": "In your case people write tickets somewhere."
  },
  {
    "start": 2259.64,
    "end": 2261.84,
    "text": "Is that GitHub issues, or is it email?"
  },
  {
    "start": 2262.04,
    "end": 2262.86,
    "text": "Or is it multi-channel."
  },
  {
    "start": 2262.9,
    "end": 2262.98,
    "text": "It's"
  },
  {
    "start": 2263.04,
    "end": 2264.2,
    "text": "basically a multi channel."
  },
  {
    "start": 2264.6,
    "end": 2270.34,
    "text": "but the AI automation we have in GitHub and our in-app support chat."
  },
  {
    "start": 2270.48,
    "end": 2276.24,
    "text": "so if you log into lengthus.com there's this small support button."
  },
  {
    "start": 2276.34,
    "end": 2281.52,
    "text": "You can hit it And then your textbox where all agent works."
  },
  {
    "start": 2281.8,
    "end": 2285.62,
    "text": "Then also hit a small button like talk to an engineer."
  },
  {
    "start": 2286.18,
    "end": 2287.78,
    "text": "So yeah This how I do."
  },
  {
    "start": 2288.6,
    "end": 2289.36,
    "text": "Okay, okay."
  },
  {
    "start": 2289.4,
    "end": 2292.44,
    "text": "And that's all proprietary self-build even open source."
  },
  {
    "start": 2292.68,
    "end": 2292.8,
    "text": "or"
  },
  {
    "start": 2294.6,
    "end": 2295.74,
    "text": "we use a vendor for this."
  },
  {
    "start": 2296.14,
    "end": 2298.92,
    "text": "but again We made sure that this vendor has all the context."
  },
  {
    "start": 2299.26,
    "end": 2301.22,
    "text": "This is where Open Source is extremely helpful."
  },
  {
    "start": 2301.86,
    "end": 2304.9,
    "text": "Our code is open sources like it accessible by an agent."
  },
  {
    "start": 2305.4,
    "end": 2309.92,
    "text": "All the documentation Everything...all the github issues are public."
  },
  {
    "start": 2310.38,
    "end": 2312.64,
    "text": "So this how quality of this agent."
  },
  {
    "start": 2313.04,
    "end": 2324.64,
    "text": "really good because you can find lots of context about bugs that we have features, how an upgrade path for a certain API looks like."
  },
  {
    "start": 2325.02,
    "end": 2330.2,
    "text": "These are all fully documented and it's out there so easy to produce a good answer."
  },
  {
    "start": 2332.2,
    "end": 2335.18,
    "text": "from your view what is the pattern right now?"
  },
  {
    "start": 2335.26,
    "end": 2339.54,
    "text": "you see often I mean many customers basically try to implement agents."
  },
  {
    "start": 2341.22,
    "end": 2346.42,
    "text": "every BtoB SaaS could be customer really shipping?"
  },
  {
    "start": 2346.56,
    "end": 2348.14,
    "text": "And what's the pattern?"
  },
  {
    "start": 2348.26,
    "end": 2351.52,
    "text": "teams keep failing or keep trying and then, than keep failing."
  },
  {
    "start": 2352.2,
    "end": 2354.7,
    "text": "Like What are they mistakes that people often make?"
  },
  {
    "start": 2355.16,
    "end": 2357.68,
    "text": "this is Really hard to answer."
  },
  {
    "start": 2358.14,
    "end": 2362.26,
    "text": "I don't have a global view on this but yeah The team said succeed."
  },
  {
    "start": 2362.84,
    "end": 2364.0,
    "text": "They just try."
  },
  {
    "start": 2364.86,
    "end": 2367.46,
    "text": "They just tried most powerful model day."
  },
  {
    "start": 2367.52,
    "end": 2371.24,
    "text": "Just give it as I just set as much context as possible."
  },
  {
    "start": 2371.64,
    "end": 2373.02,
    "text": "They try narrow use cases."
  },
  {
    "start": 2373.14,
    "end": 2378.1,
    "text": "Don't try too solve everything at once, but go step by step and then try to expand from there."
  },
  {
    "start": 2378.6,
    "end": 2383.68,
    "text": "These are the successful cases where I mean i know from some customers."
  },
  {
    "start": 2383.8,
    "end": 2390.86,
    "text": "they really understand the workflows They tried to automate like their for example founders of a YC startup."
  },
  {
    "start": 2391.08,
    "end": 2393.38,
    "text": "And they're building something For example in recruiting?"
  },
  {
    "start": 2393.74,
    "end": 2394.96,
    "text": "There's no idea about recruiting."
  },
  {
    "start": 2395.0,
    "end": 2400.88,
    "text": "that could just say hey scrape these profiles and Then give us summary and rank them somehow and then return it to The user."
  },
  {
    "start": 2401.12,
    "end": 2401.28,
    "text": "done."
  },
  {
    "start": 2401.7,
    "end": 2403.22,
    "text": "this could be shipped on a Saturday."
  },
  {
    "start": 2403.8,
    "end": 2408.92,
    "text": "But now they really try to understand how recruiting function, world-class recruiting functions works."
  },
  {
    "start": 2409.04,
    "end": 2418.36,
    "text": "And then tried to replicate this within their agent and build very thoughtful skills for agents... ...to really replicate basically what humans do!"
  },
  {
    "start": 2419.56,
    "end": 2427.72,
    "text": "They're in love with the product and trying to improve each & every nuance by doing whatever like improving skill, prompts or context."
  },
  {
    "start": 2428.2,
    "end": 2436.26,
    "text": "So there's still I think that teams succeed are the ones in the end that care a lot about quality of output and do whatever it takes to get there."
  },
  {
    "start": 2436.88,
    "end": 2441.86,
    "text": "Because you can also just ask Claude, who built your world class hiring agent right?"
  },
  {
    "start": 2441.9,
    "end": 2441.96,
    "text": "Yes"
  },
  {
    "start": 2442.42,
    "end": 2445.98,
    "text": "probably this will become better over time."
  },
  {
    "start": 2446.34,
    "end": 2452.42,
    "text": "but then again... That's an interesting question where we'll end up because it is agents building agents."
  },
  {
    "start": 2452.86,
    "end": 2453.76,
    "text": "This was also our future."
  },
  {
    "start": 2454.04,
    "end": 2456.48,
    "text": "thinking what role human?"
  },
  {
    "start": 2456.58,
    "end": 2466.84,
    "text": "at some point I currently think email slash linear inbox where an agent tells me, hey Max we needed to fix this here because it's broke."
  },
  {
    "start": 2467.1,
    "end": 2470.08,
    "text": "And then if I want too i can just dive into and see all the context."
  },
  {
    "start": 2470.2,
    "end": 2487.9,
    "text": "within this notification Agents will notify us that they did something and let us help or not help rather than us needing to launch the agent."
  },
  {
    "start": 2488.1,
    "end": 2491.14,
    "text": "And hand-hold it over the line, yeah this would be very exciting."
  },
  {
    "start": 2491.68,
    "end": 2492.72,
    "text": "So a decision a day."
  },
  {
    "start": 2492.84,
    "end": 2495.6,
    "text": "basically then ultimately right."
  },
  {
    "start": 2495.64,
    "end": 2495.74,
    "text": "here's"
  },
  {
    "start": 2495.8,
    "end": 2502.2,
    "text": "my yes your decision in box wake up in the morning at am drink of coffee go through all the decisions and Let It Go."
  },
  {
    "start": 2503.04,
    "end": 2509.74,
    "text": "and Then like if coding also Become so much faster and it's basically a parallelism right?"
  },
  {
    "start": 2509.8,
    "end": 2515.18,
    "text": "So you can just like work on hundred tasks in parallel, And then we're bound by compute from open AI."
  },
  {
    "start": 2515.78,
    "end": 2519.5,
    "text": "By the amount of this is like basically attention for human to make decisions."
  },
  {
    "start": 2520.1,
    "end": 2522.14,
    "text": "if the agent says I want have a decision!"
  },
  {
    "start": 2522.8,
    "end": 2529.18,
    "text": "Then yeah i'm wondering where we end up with because the iteration speed will be crazy fast."
  },
  {
    "start": 2529.46,
    "end": 2538.66,
    "text": "Where it used to take teams Build features now things will just happen overnight and yeah build very exciting products."
  },
  {
    "start": 2539.22,
    "end": 2541.34,
    "text": "Okay, And if you're honest?"
  },
  {
    "start": 2541.76,
    "end": 2541.94,
    "text": "Yeah"
  },
  {
    "start": 2541.98,
    "end": 2543.1,
    "text": "Where does this also suck?"
  },
  {
    "start": 2543.58,
    "end": 2545.94,
    "text": "like where there's the current status course to start?"
  },
  {
    "start": 2546.34,
    "end": 2550.06,
    "text": "Like what do your... What is that stuff really annoys me these days?"
  },
  {
    "start": 2550.38,
    "end": 2556.72,
    "text": "Is it often changing directions and new stuff around the globe every day?"
  },
  {
    "start": 2556.96,
    "end": 2563.14,
    "text": "or I don't know, the fact that everyone now releases agent frameworks and like?"
  },
  {
    "start": 2563.68,
    "end": 2568.18,
    "text": "is it the non-existence of product market fit these days?"
  },
  {
    "start": 2568.7,
    "end": 2575.08,
    "text": "Or are there really technical issues where you think hey this too basic we're all cooking with water."
  },
  {
    "start": 2575.88,
    "end": 2592.9,
    "text": "What does it... It's bad at same time as not a clear path on what to be doing in an engineering organization should built lots of infrastructure to make sure that our coding agents are able to code without reading from our production database and destroying stuff."
  },
  {
    "start": 2592.98,
    "end": 2597.94,
    "text": "So there needs to be a lot of layers around arbitrary code execution by an agent, for example."
  },
  {
    "start": 2599.04,
    "end": 2603.5,
    "text": "Should we now go and invest heavily into this, build all these sort of things?"
  },
  {
    "start": 2603.66,
    "end": 2605.36,
    "text": "Or should we wait for someone to build it?"
  },
  {
    "start": 2606.28,
    "end": 2607.98,
    "text": "or will something else be the future."
  },
  {
    "start": 2608.48,
    "end": 2621.68,
    "text": "I'm fairly convinced that we'll do this but sometimes you don't know what's best next step because people are constrained."
  },
  {
    "start": 2621.74,
    "end": 2626.24,
    "text": "so making the right bets is probably the biggest challenge."
  },
  {
    "start": 2627.92,
    "end": 2631.3,
    "text": "So really like figuring out what you don't know, right?"
  },
  {
    "start": 2631.38,
    "end": 2631.6,
    "text": "Like."
  },
  {
    "start": 2632.84,
    "end": 2636.22,
    "text": "What do people maybe are doing in the background that we don't about?"
  },
  {
    "start": 2636.56,
    "end": 2638.96,
    "text": "where does it actually open?"
  },
  {
    "start": 2639.06,
    "end": 2647.4,
    "text": "source access for this case is opening up working on this or whatever I can imagine at this rate and there's also a problem of many."
  },
  {
    "start": 2649.88,
    "end": 2652.5,
    "text": "Basically start doing something and then realize, okay shit."
  },
  {
    "start": 2652.74,
    "end": 2653.72,
    "text": "This has been built like."
  },
  {
    "start": 2654.08,
    "end": 2654.9,
    "text": "I don't need to build this."
  },
  {
    "start": 2655.62,
    "end": 2659.7,
    "text": "It happens every day And you throw away two thirds of the stuff your building right?"
  },
  {
    "start": 2660.14,
    "end": 2662.52,
    "text": "Yeah But if you build it You can customize."
  },
  {
    "start": 2662.76,
    "end": 2664.26,
    "text": "that's always a nice thing."
  },
  {
    "start": 2664.56,
    "end": 2668.68,
    "text": "If you build stuff in-house... ...you can make it work for use case."
  },
  {
    "start": 2669.46,
    "end": 2681.52,
    "text": "Also what i believe will happen is really Big Tech have done it in US Building their own tooling internally like their internal task tracking and email clients, whatnot."
  },
  {
    "start": 2681.98,
    "end": 2690.34,
    "text": "And I think this will happen to way more companies to build custom internal tooling for your workflows because building software would become cheaper."
  },
  {
    "start": 2690.58,
    "end": 2694.38,
    "text": "then you can adjust it the workflow that you care about."
  },
  {
    "start": 2695.98,
    "end": 2708.44,
    "text": "one little case study what i have for you in a real life case study maybe slightly I'm slightly lying but parts of it are real and i'll tell you which ones."
  },
  {
    "start": 2708.98,
    "end": 2715.36,
    "text": "Like, i'm co-founder of sas group And we acquire roughly like thirty portfolio companies that belong to us really?"
  },
  {
    "start": 2715.46,
    "end": 2717.72,
    "text": "Like end-to-end tech billing etc."
  },
  {
    "start": 2718.28,
    "end": 2719.64,
    "text": "all Of that is owned by us."
  },
  {
    "start": 2720.12,
    "end": 2720.92,
    "text": "We have support."
  },
  {
    "start": 2721.0,
    "end": 2724.72,
    "text": "we have on call own ops setups individual teams."
  },
  {
    "start": 2725.54,
    "end": 2730.7,
    "text": "people partly use aws partly used google cloud parties datalog."
  },
  {
    "start": 2730.8,
    "end": 2736.26,
    "text": "so and like the old world, a bit of an IT nightmare security nightmare."
  },
  {
    "start": 2736.78,
    "end": 2739.28,
    "text": "And very fragmented."
  },
  {
    "start": 2740.0,
    "end": 2743.84,
    "text": "if you see this all and zoomed out right?"
  },
  {
    "start": 2744.3,
    "end": 2745.06,
    "text": "You were in my shoes."
  },
  {
    "start": 2745.66,
    "end": 2748.74,
    "text": "what would be first thing that you try to solve with agents?"
  },
  {
    "start": 2749.72,
    "end": 2750.54,
    "text": "where will it start?"
  },
  {
    "start": 2750.62,
    "end": 2752.86,
    "text": "how would the infrastructure look like?"
  },
  {
    "start": 2753.42,
    "end": 2764.36,
    "text": "wouldn't it be on-call setup support because we could generate most value using length use And what would be like the concrete stuff you would do?"
  },
  {
    "start": 2764.84,
    "end": 2769.06,
    "text": "Yeah, it's all different businesses that are... Different teams etc."
  },
  {
    "start": 2769.38,
    "end": 2770.82,
    "text": "Everybody working on their own products."
  },
  {
    "start": 2771.22,
    "end": 2771.32,
    "text": "yeah"
  },
  {
    "start": 2772.28,
    "end": 2775.34,
    "text": "Yes yes and a few things sort of uncommon."
  },
  {
    "start": 2775.42,
    "end": 2778.66,
    "text": "Like billing is often done through Stripe support."
  },
  {
    "start": 2778.76,
    "end": 2781.8,
    "text": "It's often done for one solution or other."
  },
  {
    "start": 2782.48,
    "end": 2788.46,
    "text": "code is on GitHub, but you don't have like a context described necessarily per company."
  },
  {
    "start": 2788.58,
    "end": 2790.28,
    "text": "You do not have MCP servers per company."
  },
  {
    "start": 2790.82,
    "end": 2805.58,
    "text": "would be like MCPing everything and then trying to really define the context for companies or basically approach the on-call stuff that you describe security what it will be?"
  },
  {
    "start": 2806.8,
    "end": 2810.04,
    "text": "Yeah, if you don't want to merge these businesses and make them one."
  },
  {
    "start": 2810.48,
    "end": 2814.32,
    "text": "You probably want to have individual setups for each of these companies."
  },
  {
    "start": 2814.8,
    "end": 2821.3,
    "text": "basically makes sure everybody's AI built And thinks about their job Each and every day Of how they can use AI To make it better."
  },
  {
    "start": 2822.2,
    "end": 2824.8,
    "text": "Like there are so many things like in recruiting."
  },
  {
    "start": 2825.26,
    "end": 2828.28,
    "text": "We also did that cloud co-work to improve the recruiting process."
  },
  {
    "start": 2828.42,
    "end": 2838.26,
    "text": "And then maybe you have some shared learnings where one recruiting tool works better with AI than another one, and then you can share this knowledge across different companies that you have of like tools selection."
  },
  {
    "start": 2838.66,
    "end": 2839.78,
    "text": "but it will also change a lot."
  },
  {
    "start": 2839.96,
    "end": 2846.38,
    "text": "people still need to understand all the tools they are using because we're not there yet just on autopilot."
  },
  {
    "start": 2848.48,
    "end": 2856.4,
    "text": "so I would say AI pulled the workforce, everybody that they start to use these tools for their internal processes."
  },
  {
    "start": 2856.86,
    "end": 2863.0,
    "text": "And then I would heavily think about a product... Right now most or all of this products have power market fit."
  },
  {
    "start": 2863.44,
    "end": 2865.18,
    "text": "so understand where do the products move?"
  },
  {
    "start": 2865.52,
    "end": 2871.48,
    "text": "and maybe you need something more agentic because it could also lose."
  },
  {
    "start": 2871.9,
    "end": 2877.36,
    "text": "uh, product market fit at some point because whatever your users do right now with their tools may not work in the future."
  },
  {
    "start": 2877.42,
    "end": 2879.42,
    "text": "So what's the feature of all this?"
  },
  {
    "start": 2879.48,
    "end": 2880.5,
    "text": "In a more authentic world?"
  },
  {
    "start": 2881.04,
    "end": 2882.06,
    "text": "so I would prepare for that."
  },
  {
    "start": 2882.56,
    "end": 2887.34,
    "text": "okay Okay and um i think i mean That's what we do."
  },
  {
    "start": 2888.14,
    "end": 2897.46,
    "text": "There are some things that you could still provide to every company like i mean having those individual teams by it Having individual engineers individuals support people etc."
  },
  {
    "start": 2898.24,
    "end": 2905.76,
    "text": "there's Still something some some central baseline that you can draw, and I think like on-call for example would be something."
  },
  {
    "start": 2905.84,
    "end": 2915.44,
    "text": "or this let's say knowledge about building an agentic one call process right is something that you Can infuse also?"
  },
  {
    "start": 2915.7,
    "end": 2919.34,
    "text": "Like really raising the bar to a certain extent when it comes."
  },
  {
    "start": 2919.42,
    "end": 2921.44,
    "text": "so like SLIs SLOs etc."
  },
  {
    "start": 2922.7,
    "end": 2925.04,
    "text": "Observability as something that You can do Really."
  },
  {
    "start": 2926.08,
    "end": 2927.18,
    "text": "And then support Etc."
  },
  {
    "start": 2928.0,
    "end": 2929.8,
    "text": "I think there are agent cases, right?"
  },
  {
    "start": 2929.94,
    "end": 2933.26,
    "text": "And and that's also the translation problem between like central people."
  },
  {
    "start": 2933.3,
    "end": 2936.82,
    "text": "That work for this central organization in the individual organizations."
  },
  {
    "start": 2937.16,
    "end": 2937.16,
    "text": "I"
  },
  {
    "start": 2945.3,
    "end": 2946.74,
    "text": "mean, there's two pieces like."
  },
  {
    "start": 2947.02,
    "end": 2948.48,
    "text": "one thing is to work flows."
  },
  {
    "start": 2948.64,
    "end": 2950.52,
    "text": "Like how do you stitch tools together?"
  },
  {
    "start": 2950.7,
    "end": 2951.94,
    "text": "So they work well."
  },
  {
    "start": 2952.04,
    "end": 2962.02,
    "text": "in an agentic world i think this Is mostly easy because uh... There are so many tools that help You stitch together tools or you can build stuff in house."
  },
  {
    "start": 2962.06,
    "end": 2965.58,
    "text": "so stitching Together tools should be rather simple."
  },
  {
    "start": 2966.0,
    "end": 2970.3,
    "text": "and then the big question about Agent MD in your codebase."
  },
  {
    "start": 2970.76,
    "end": 2972.34,
    "text": "How high quality is the agent md?"
  },
  {
    "start": 2972.58,
    "end": 2978.94,
    "text": "In a code base and it should, It needs to be super custom To your application what are the common error scenarios?"
  },
  {
    "start": 2979.1,
    "end": 2981.08,
    "text": "What are the edge cases that users run into?"
  },
  {
    "start": 2981.18,
    "end": 2982.46,
    "text": "where do you have scaling problems?"
  },
  {
    "start": 2982.94,
    "end": 2984.24,
    "text": "What does your design system?"
  },
  {
    "start": 2984.36,
    "end": 2986.52,
    "text": "if You build a new feature how should it look like?"
  },
  {
    "start": 2986.78,
    "end": 2990.2,
    "text": "this will be very different for all The different products under different companies."
  },
  {
    "start": 2990.64,
    "end": 2999.3,
    "text": "thus I would say these pieces need to be highly specific for these applications, how the product works."
  },
  {
    "start": 3000.54,
    "end": 3005.56,
    "text": "This needs very specific and people working on this need to maintain it over time."
  },
  {
    "start": 3006.16,
    "end": 3008.42,
    "text": "Okay so that's like a normal desk workflow."
  },
  {
    "start": 3008.64,
    "end": 3013.16,
    "text": "you also recommend improving everyone out there right?"
  },
  {
    "start": 3013.22,
    "end": 3020.02,
    "text": "Start collaborating with Claude in Codex like pushing the context, start optimizing it."
  },
  {
    "start": 3020.12,
    "end": 3023.42,
    "text": "Start sharing it and then really start continuing to develop it."
  },
  {
    "start": 3023.52,
    "end": 3029.28,
    "text": "so make a workflow another one of things that ends up on your individual hard drive right?"
  },
  {
    "start": 3031.44,
    "end": 3041.48,
    "text": "And what's you view on share or required minimum level best practice when it comes to multi-agent workflows etc."
  },
  {
    "start": 3042.28,
    "end": 3050.58,
    "text": "is there something where personally do use Like an engineering team on your hard disk or like a virtual one."
  },
  {
    "start": 3050.78,
    "end": 3053.06,
    "text": "Yeah, are you using multiple roles?"
  },
  {
    "start": 3053.2,
    "end": 3059.18,
    "text": "Are you doing like cloud code and step-by-step Or how are you personally doing that?"
  },
  {
    "start": 3059.66,
    "end": 3067.3,
    "text": "so we have if We have a bunch of skills for different things in the company And then basically selecting the right skill with the with the harness to do something."
  },
  {
    "start": 3067.34,
    "end": 3073.04,
    "text": "So it's currently mostly differentiated for tasks by skills, and giving it one off-context."
  },
  {
    "start": 3073.18,
    "end": 3075.68,
    "text": "So we were working on SOC II and TIPAA"
  },
  {
    "start": 3075.82,
    "end": 3075.9,
    "text": "etc.,"
  },
  {
    "start": 3076.74,
    "end": 3080.48,
    "text": "just again the right context with all of our tasks."
  },
  {
    "start": 3080.84,
    "end": 3089.3,
    "text": "so I try to run things in parallel but don't have one agent orchestrating many agents like that running now."
  },
  {
    "start": 3090.5,
    "end": 3092.02,
    "text": "Do you think this would be a big step?"
  },
  {
    "start": 3092.4,
    "end": 3095.94,
    "text": "still if you add or are not curious?"
  },
  {
    "start": 3097.28,
    "end": 3098.46,
    "text": "I'd be curious to try it out."
  },
  {
    "start": 3098.82,
    "end": 3113.78,
    "text": "Right now, i think we need improve the loop of a single agent To make this better because then like an one-agent Controlling many agents would just launch different tasks Then instead of one task have twenty tasks to review."
  },
  {
    "start": 3114.52,
    "end": 3127.44,
    "text": "I'm still the bottleneck at reviewing and basically merging stuff or sending out this email doing whatever I told the agent to do, so basically improving this piece of the loop is probably higher leverage right now."
  },
  {
    "start": 3128.5,
    "end": 3139.26,
    "text": "Okay okay yeah fair enough i think that's actually the status quo um of many companies out there and i guess you're a bit further ahead than most."
  },
  {
    "start": 3139.32,
    "end": 3141.78,
    "text": "uh because you also building AI native solutions."
  },
  {
    "start": 3142.82,
    "end": 3158.02,
    "text": "but i also see many people who try really go step by step still skills for the stuff that they need to kind of accelerate their software development lifecycle, right?"
  },
  {
    "start": 3158.12,
    "end": 3162.76,
    "text": "I mean do you believe in things like pull requests being alive on a few years still?"
  },
  {
    "start": 3163.04,
    "end": 3166.14,
    "text": "or Do you think this will won't be required at some point."
  },
  {
    "start": 3166.8,
    "end": 3171.24,
    "text": "I mean a pull request is yes Yes But how fast it would be?"
  },
  {
    "start": 3171.34,
    "end": 3173.0,
    "text": "merged will be much faster."
  },
  {
    "start": 3173.14,
    "end": 3175.54,
    "text": "i think there will be interesting topics."
  },
  {
    "start": 3176.48,
    "end": 3185.66,
    "text": "The load under GitHub systems is incredibly high by now, mostly because like so many pull requests are opened and then they run the GitHub actions over-and-over."
  },
  {
    "start": 3186.16,
    "end": 3188.36,
    "text": "And now we're opening some new pull request."
  },
  {
    "start": 3188.48,
    "end": 3191.38,
    "text": "GitHub Actions Run... ...and make sure that this system still works!"
  },
  {
    "start": 3191.42,
    "end": 3208.94,
    "text": "So I think having a system where you have this Diffing and understanding whether the New thing works better than The Old Thing... This functionality is incredibly important and also good checkpoint for an agent to see what's the diff what's the incremental diff of the pull request."
  },
  {
    "start": 3209.32,
    "end": 3213.58,
    "text": "If something has again changed on a pull request, see whether all tests succeed and then merge?"
  },
  {
    "start": 3214.08,
    "end": 3226.0,
    "text": "And there will be problems such as merge queues where now ten pull requests wait to be merged Where lots of work needs to go into CI pipelines again To make sure they're not running for half an hour but less than ten minutes."
  },
  {
    "start": 3226.62,
    "end": 3231.18,
    "text": "So yeah it'll be ddsci pipelines Will there."
  },
  {
    "start": 3231.22,
    "end": 3236.72,
    "text": "they will need to be a lot of work to make them perform and then fast so we can iterate fast because at some point this would be the bottleneck."
  },
  {
    "start": 3237.24,
    "end": 3240.44,
    "text": "This will build the bottlenecks right from engineering capacity, too?"
  },
  {
    "start": 3240.48,
    "end": 3246.78,
    "text": "Like them whole rest off the workflow an and then ultimately to decision yes certain points as we"
  },
  {
    "start": 3247.1,
    "end": 3251.0,
    "text": "decisions tastes and then merging stuff and get it live."
  },
  {
    "start": 3251.82,
    "end": 3254.32,
    "text": "Do you think what's your perspective on UIs?"
  },
  {
    "start": 3254.42,
    "end": 3265.32,
    "text": "like I mean looking at our portfolio I want to encourage people, maybe step back a little and don't be in love with the UI."
  },
  {
    "start": 3265.5,
    "end": 3268.24,
    "text": "Do you think this is gonna come?"
  },
  {
    "start": 3268.36,
    "end": 3271.0,
    "text": "Like do you think there's future without your eyes or..."
  },
  {
    "start": 3271.62,
    "end": 3283.14,
    "text": "Heavily debated our latest road mapping discussions because if i look at our roadmap we built traditional UI features such as make filters better for users to navigate quickly."
  },
  {
    "start": 3283.74,
    "end": 3290.28,
    "text": "And then we also have this in-app agent, which should be able to render custom one off UI for customers."
  },
  {
    "start": 3291.08,
    "end": 3293.54,
    "text": "and Then we have the discussion round MCP servers."
  },
  {
    "start": 3293.64,
    "end": 3294.58,
    "text": "Make everything headless."
  },
  {
    "start": 3294.66,
    "end": 3297.16,
    "text": "basically just use APIs for Everything."
  },
  {
    "start": 3297.64,
    "end": 3302.96,
    "text": "and then let's The ones who just want to navigate length views with skills and cloud code."
  },
  {
    "start": 3303.46,
    "end": 3306.5,
    "text": "Let them succeed and make them never lock into the product."
  },
  {
    "start": 3307.06,
    "end": 3314.06,
    "text": "and I think Kind of everything off this needs to happen because it just let's go through the user journey."
  },
  {
    "start": 3314.38,
    "end": 3316.44,
    "text": "Someone wants to start using length use they want."
  },
  {
    "start": 3316.6,
    "end": 3323.12,
    "text": "don't want to fiddle around in the UI copy API keys and then read a documentation on how do you install the SDK?"
  },
  {
    "start": 3323.76,
    "end": 3328.82,
    "text": "No, MCP server, sorry skill MCP Server."
  },
  {
    "start": 3329.18,
    "end": 3334.58,
    "text": "and then just let Claude code work until all the data is correctly tracked with a length use."
  },
  {
    "start": 3334.66,
    "end": 3335.06,
    "text": "And tracing."
  },
  {
    "start": 3335.12,
    "end": 3339.3,
    "text": "it's currently set up... ...and give to user just link to product of like."
  },
  {
    "start": 3339.34,
    "end": 3341.08,
    "text": "hey here you can see your test traces."
  },
  {
    "start": 3341.5,
    "end": 3342.4,
    "text": "do you think that makes sense?"
  },
  {
    "start": 3342.98,
    "end": 3347.3,
    "text": "And here are the cost and all metadata that is already ingested."
  },
  {
    "start": 3347.76,
    "end": 3352.74,
    "text": "Then giving a custom link to user who can hit it, then go in a pretty UI."
  },
  {
    "start": 3353.14,
    "end": 3356.1,
    "text": "this could be one off or like general UI."
  },
  {
    "start": 3356.62,
    "end": 3362.84,
    "text": "so the UI piece still important because Also, maybe users don't spend that much time in UI anymore."
  },
  {
    "start": 3363.32,
    "end": 3369.98,
    "text": "So then having a very intuitive UI this easy to navigate will still be super important."
  },
  {
    "start": 3370.52,
    "end": 3383.1,
    "text": "and the last piece is around agentic workflows within Langfuse where you could for example say I want create synthetic data sets with Langfus like basically test sets."
  },
  {
    "start": 3383.14,
    "end": 3394.88,
    "text": "This can be complex workflow ish UI, or you have an agent proposing these based on some questionnaire that may render one off for you."
  },
  {
    "start": 3395.4,
    "end": 3406.4,
    "text": "And then it's... You just accept stuff in a ... It's more than like a chat UI with custom components rendered into the chats and yeah That's basically how we envisioned right now."
  },
  {
    "start": 3407.08,
    "end": 3411.46,
    "text": "Okay Yeah cool This is also how I envisioned it."
  },
  {
    "start": 3411.7,
    "end": 3419.08,
    "text": "I think it was Gregory Ortrick recently posting about design now becoming a super important skill."
  },
  {
    "start": 3419.62,
    "end": 3428.48,
    "text": "And I agree for those small, let's say micro UIs that you kind of get in the middle of chatting to an agent."
  },
  {
    "start": 3429.82,
    "end": 3439.36,
    "text": "but he mentioned Airbnb as a design product which people love using and I slightly disagree."
  },
  {
    "start": 3439.96,
    "end": 3441.12,
    "text": "I think, yes."
  },
  {
    "start": 3441.48,
    "end": 3447.2,
    "text": "I like Airbnb and the platform or the venues that i find there right?"
  },
  {
    "start": 3447.74,
    "end": 3453.9,
    "text": "And it comes with everything you need around to find a right venue."
  },
  {
    "start": 3454.16,
    "end": 3457.2,
    "text": "but let's say if we can automate this process then don't look at Airbnb..."
  },
  {
    "start": 3457.66,
    "end": 3458.06,
    "text": "Yeah!"
  },
  {
    "start": 3458.12,
    "end": 3458.68,
    "text": "...I'm also happy."
  },
  {
    "start": 3459.58,
    "end": 3473.54,
    "text": "You may need to book your train of flight or rental car alongside and would be great for one UI on an entire trip to see what you're actually planning there and how everything matches in terms of timelines."
  },
  {
    "start": 3474.18,
    "end": 3484.74,
    "text": "So yeah, all these tools would need some API MCPs set up And you will have the capability build quick front ends one off for case that your interested In."
  },
  {
    "start": 3485.64,
    "end": 3497.74,
    "text": "You basically want a journey registry where all the MCP servers are in All this CLI tools which then basically books the train books, the venue etc."
  },
  {
    "start": 3498.0,
    "end": 3506.18,
    "text": "But I think for many companies it's super painful to and i can't imagine that if i was Airbnb For me It would also be painful To say hey!"
  },
  {
    "start": 3506.46,
    "end": 3509.72,
    "text": "I let go this Just for a better tomorrow."
  },
  {
    "start": 3510.38,
    "end": 3514.8,
    "text": "It's painful...I called myself A design company but maybe now its no longer That important."
  },
  {
    "start": 3515.02,
    "end": 3525.5,
    "text": "And uh..i want to give people still The best presides & the best venues possible And thats my job & necessarily um..the best UI around it at the best possible search."
  },
  {
    "start": 3525.78,
    "end": 3525.94,
    "text": "right"
  },
  {
    "start": 3526.62,
    "end": 3529.86,
    "text": "i think yes and this is how I think companies will change in."
  },
  {
    "start": 3529.96,
    "end": 3532.54,
    "text": "we'll have to change because just everything."
  },
  {
    "start": 3532.92,
    "end": 3534.38,
    "text": "Will change of."
  },
  {
    "start": 3534.46,
    "end": 3542.98,
    "text": "like digital products, and I totally agree with you can't try to force your users to use a platform you intend them to use."
  },
  {
    "start": 3543.02,
    "end": 3545.04,
    "text": "that let then do their thing."
  },
  {
    "start": 3545.7,
    "end": 3548.06,
    "text": "maybe were also bit Marty, right?"
  },
  {
    "start": 3548.48,
    "end": 3551.64,
    "text": "I mean we're deep down."
  },
  {
    "start": 3552.16,
    "end": 3559.1,
    "text": "And many people still maybe for your girlfriend it would be different and she values the experience."
  },
  {
    "start": 3559.2,
    "end": 3573.94,
    "text": "on really searching on Airbnb doesn't love chatting to Claude so this will a process I assume at least like...I can imagine what but where this won't be case in year."
  },
  {
    "start": 3574.5,
    "end": 3575.52,
    "text": "Yes, agree."
  },
  {
    "start": 3575.6,
    "end": 3589.54,
    "text": "Like there will always be some default UI that people enjoy and love because it's also a brand and is thoughtful and well used but again like closing against the idea and shutting down the idea of being more open about it."
  },
  {
    "start": 3590.02,
    "end": 3593.12,
    "text": "I think this was not the right way so probably needs to exist both in parallel."
  },
  {
    "start": 3593.7,
    "end": 3595.46,
    "text": "okay agreed."
  },
  {
    "start": 3595.52,
    "end": 3600.899,
    "text": "then Before we come to the end, one maybe recommendation from your end."
  },
  {
    "start": 3602.0,
    "end": 3610.759,
    "text": "I assume you treated quite similar like agent kit that Vercel released with some open examples?"
  },
  {
    "start": 3611.24,
    "end": 3619.22,
    "text": "Do have good example where if people want a dig deeper they can check later on To basically start with using your product."
  },
  {
    "start": 3619.299,
    "end": 3620.46,
    "text": "I guess you have right."
  },
  {
    "start": 3620.56,
    "end": 3621.68,
    "text": "like what's the best example?"
  },
  {
    "start": 3622.48,
    "end": 3628.279,
    "text": "You can give to people that they, uh... That i can link in the show notes and people can check them."
  },
  {
    "start": 3628.339,
    "end": 3639.839,
    "text": "I think what we do is basically you can just go to lengthviews.com slash demo and there's a small chatbot that's on top of all our contexts, You can ask how to use Lengthviews?"
  },
  {
    "start": 3640.22,
    "end": 3641.1,
    "text": "Why should i use it?"
  },
  {
    "start": 3641.18,
    "end": 3641.259,
    "text": "etc."
  },
  {
    "start": 3642.06,
    "end": 3649.2,
    "text": "And then there's the small button login project where you can see what happened."
  },
  {
    "start": 3649.299,
    "end": 3651.66,
    "text": "So, You Can See the Traces of The Chatbot You Just Used?"
  },
  {
    "start": 3651.779,
    "end": 3657.02,
    "text": "You Can see LLM as a Judge Evaluating Whether The Answer Was Good or Not Of This Chatbot."
  },
  {
    "start": 3657.6,
    "end": 3661.839,
    "text": "It's A Very Easy Way To Interact With The Product And See What's Going On There."
  },
  {
    "start": 3662.42,
    "end": 3664.56,
    "text": "Okay so try them all quite easy."
  },
  {
    "start": 3665.62,
    "end": 3670.12,
    "text": "and then a little surprise from my end I spoke to your course owner Mark Lingen."
  },
  {
    "start": 3671.7,
    "end": 3674.379,
    "text": "He told me about the hidden Easter egg and length use."
  },
  {
    "start": 3674.879,
    "end": 3686.02,
    "text": "So I don't know if you knew that, but if your created trace with a timestamp from before language languages existed this system actually pulls you to that very moment in time."
  },
  {
    "start": 3686.62,
    "end": 3695.58,
    "text": "And we now imagine That we go back for the year twenty fourteen when he was still studying at teo munica thing."
  },
  {
    "start": 3695.74,
    "end": 3697.24,
    "text": "and you're building iOS apps?"
  },
  {
    "start": 3698.16,
    "end": 3702.56,
    "text": "You can't imagine what what might have But what might happen in a few years?"
  },
  {
    "start": 3702.66,
    "end": 3707.799,
    "text": "Like, What would be the one sentence that you would whisper into your younger self's ears."
  },
  {
    "start": 3708.64,
    "end": 3710.24,
    "text": "Be curious and work hard!"
  },
  {
    "start": 3713.56,
    "end": 3717.94,
    "text": "Okay but I think this is actually what we did"
  },
  {
    "start": 3717.98,
    "end": 3718.54,
    "text": "right."
  },
  {
    "start": 3718.58,
    "end": 3718.7,
    "text": "so... You"
  },
  {
    "start": 3718.74,
    "end": 3719.54,
    "text": "don't change anything"
  },
  {
    "start": 3719.72,
    "end": 3723.839,
    "text": "No And enjoy the ride It was so exciting."
  },
  {
    "start": 3724.62,
    "end": 3727.0,
    "text": "and buy bitcoins so that you don't have to sell your company."
  },
  {
    "start": 3730.06,
    "end": 3730.92,
    "text": "No investment advice here!"
  },
  {
    "start": 3731.899,
    "end": 3732.58,
    "text": "Okay, cool."
  },
  {
    "start": 3733.08,
    "end": 3742.12,
    "text": "So thanks a lot Max Was great having you And I wish all the success with Langfuse Really looking forward to observe your crazy ride."
  },
  {
    "start": 3743.62,
    "end": 3744.74,
    "text": "Hope to see you soon."
  },
  {
    "start": 3745.2,
    "end": 3746.66,
    "text": "Thankyou for the time Tobias"
  },
  {
    "start": 3747.319,
    "end": 3749.52,
    "text": "Bye."
  },
  {
    "start": 3749.56,
    "end": 3751.759,
    "text": "Thank you for listening to the AlphaList podcast."
  },
  {
    "start": 3751.96,
    "end": 3755.56,
    "text": "If you liked this episode, share it with friends – I'm sure they love it too!"
  },
  {
    "start": 3755.879,
    "end": 3762.6,
    "text": "Make sure to subscribe so that you can hear deep insights into technical leadership and technology trends as they become available."
  },
  {
    "start": 3763.02,
    "end": 3770.18,
    "text": "Also please tell us if there is a topic you would like more about or a technical leader whose brain we'd like us pick."
  },
  {
    "start": 3777.839,
    "end": 3781.64,
    "text": "Please send us suggestions to cto at alphalist.com."
  },
  {
    "start": 3782.2,
    "end": 3783.96,
    "text": "Send me a message on LinkedIn or Twitter."
  },
  {
    "start": 3784.359,
    "end": 3793.06,
    "text": "After all, the more knowledge we bring to CTOs The more growth We see in tech Or as we say an Alphalest Accumulated Knowledge To Accelerate Growth."
  },
  {
    "start": 3793.62,
    "end": 3804.56,
    "text": "See you In next episode!"
  }
]