%20copy.png)
Adir Ben-Yehuda: So instead of like hiring an engineer, they will understand what's going on the code base and do it back and forth and like, oh, I forgot to ask you this or I need to do that. I did everything here before before our call and now everything is ready. And guess what? Instead of hiring a UX designer, I have a studio that understands the product itself and what you guys are doing. And I send everything to them.Sam Nadler: This Week, breaking it down. Built This Week, we show you how. A fresh idea, a clever tweak, you locked in shoe. Built This Week. Hey, everyone, and welcome to Built This Week, the podcast where we share what we're building, how we're building it, and what it means for the world of AI and startups.I'm Sam Nadler, cofounder here at Rise Labs, and each and every week, I'm joined by my friend, business partner, and cohost, Jordan Metzner. How are doing today, Jordan?Jordan Metzner: Hey, Sam. Happy to be back. Another exciting episode. Great guest this week. It's been a crazy week in AI.You know, record fundraising from Anthropic, Karpathy going Anthropic, the Elon OpenAI lawsuit ending. Just now OpenAI may be filing to go public by the end of the week. So, I mean, honestly, just another full week of crazy news. Google as well with new models and and everything they've been dropping. So and then even Cursor with their new models.So, yeah, it's just been crazy crazy week and hard to keep up to be honest. But, yeah, looking forward to today's episode and then and chatting about some new products.Sam Nadler: Yeah. Absolutely. And some of those stories you mentioned, we'll cover at the end, and I'm really excited to present our guests. But before I do, don't forget to like and subscribe. We have new episodes of Built This Week every Friday, I think, the 27,000 subscriber mark on YouTube.Thank you for your support. And with that, I'd like to introduce our guest, Adir, from Autonomy AI. If you don't mind, if you could give us just a quick intro on who you are, and then we'll jump into a demo.Adir Ben-Yehuda: Sure thing, guys. Thank you for having me here. It's great. My name is Adir, as you mentioned. I'm the CEO and cofounder of autonomy.ai.I've been in the industry for the past fifteen plus years with the background in go to market. And in the last two years, founded and running Autonomy AI. And, you know, it's it's always a crazy week in AI, isn't it?Sam Nadler: It is. Absolutely. And I'm act actually really excited to see what you're building because to be honest, I don't really know what it is. I I understand it's this organizational layer for, like, enterprise teams to begin using AI in, like, a in an organized systematic way. But I would just love to hear how does it work?What does it do? I think you're gonna show us something here, so I'm really excited.Adir Ben-Yehuda: That's awesome. Yeah. So let's start from the basics. Right? I think, like, generally and we're gonna talk later on about the news, but in general, the part of coding, the engineering velocity, I should say, has been resolved from that perspective, right, with AI.Engineers are no longer writing code, and they can do things very quickly and very fast. But the main problem for organizations is still the handoff. Right? How do you actually build an operating model that will support AI in general? And the main bottleneck across the board the main bottleneck across the board is that PMs, designers are still working in silos.PMs and designers are still the bottleneck between the handoff between two engineering. And even even if you go ahead and say, hey, PMs can use Cloud Code to actually be part of this process. What you see across the board and like many organizations are are seeing the same thing, it's very hard for them. Right? They're not technical individuals.They don't understand CLI. They don't understand the code base. Even if they contain the beast of codecs for that matter, they still need to be very prescribed about what they ask from that perspective. And what we've built in the last couple of years and launched it six months ago today with over a 160 companies and growing every week is this operating system on the existing code base. Our ability to actually understand an existing code base, it doesn't matter if it's like fresh new or legacy code base to understand what are the coding standards, and I can show you this in a second, what are effectively the the tribal knowledge.What is legacy? What is not legacy? But also understand semantically what the code base is doing. And then with a simple web app application, you can do everything you want on your on your platform. No need of engineering.You can be a product person. You could be a designer all the way to a pull request and push it into production. And that's where autonomy comes in and that's the layer we've built. And in the last six months, we're just like seeing more and more demand out of this from companies that are looking to do this first or companies that have tried different solutions, right, and like failed. I like to think about us as like if ClotCode or Codecs are MS DOS for the younger audience, MS DOS was an operating system back in the days.It was like I I gave this like metaphor to someone who was like, what? And so I'm like MS DOS was the early operating system, you know, green line, you need to put a floppy disk, which was like how you operate a computer. We're building Windows. Right? Easy, clickable interface, you don't need to think about it.So that's kind of like the the introduction to what we do. If you guys don't mind, I can jump in and kind of show you the main interface and we we could take it from there.Jordan Metzner: Yeah. Let's do it. Perfect.Sam Nadler: Building with AI shouldn't take months. We do it in weeks. Rise Labs helps companies design, build, and scale AI products with elite engineers from Latin America. Real teams, real velocity, real outcomes. This is Rise Labs.Adir Ben-Yehuda: So that's basically the homepage of Autonomy AI, the of the app for Autonomy AI. Very friendly, very like a simple interface. You know it probably from Bold or Lovable or all those kind of things. Right? You can actually enter a request.The input could also be an image, Figma, even a cloud cloud design. And then effectively, you can build everything from new or you can actually update existing screen. But and that's the the big part of it. That's a big thing. Once you start working with a product, you can actually connect your own repository.So for an organization, you can connect your own repository. Instead of working in silos, like, maybe with your design system or something like this, we analyze the entire code base itself. So for a CEO of an AI company, all of a sudden, they get this firepower of another platform that can actually understand their code base, in-depth, give them a full understanding of, like, how the coding standard looks like so they can have the validation of it. Right? Understand what components they have on the code base itself, and even render the flow or render it on the product itself.And from that moment on, hey, give me your summer intern. They can push things into production from that from that moment on. Right? So I connected our repository, our project, the Autonomy project. And effectively what I did is in in our simple interface, you can actually ask here in the chat the different requests.And here on this side, we actually render in real time the results. But everything is on your platform, on your code base end to end. So the first request I had is I asked, hey, can you find a task list page for us? A component from our product that looks like a task list. And you can see in a fully autonomous way, it added a Kanban view.And you can see the Kanban board was organized without us even asking based on those tests itself. And if I'll open it for full view, you can go ahead and see the this Kanban. So all of a sudden, have a whole new feature that I very easily can send it to dev with a full pull request that is ready. What do I mean by ready? Unlike other products in the market, we understand autonomously the entire code base.Right? So Face Studio sends everything with the connections to the back end, with the validations, with the testing of the product itself, with understanding like simple things like accessibility within the code itself. Again, you don't need to think. All you need to do is ask the request. We'll do everything on on your code.Face to you will do everything on your code. But I decided that I don't wanna finish it there. What I asked again is on the second task, asked, now can you add a Gantt view to it. Right? And I also wanna make sure that the Kanban is draggable.I didn't ask it before, but I wanna make sure it's like fancy and it's fully draggable and so on and so forth. So in the meantime, I'll I'll show you what phase two you did. It said said, okay. Great. I added a Gantt view with the time visualization of it, showing all the task itself.I made the cam button also draggable and I made sure that the list view, the original expandable list view with all test details. So you can actually click the test themselves and it will be fully extend expandable. So now from the initial list view that I had, I have a cam view, fully operational. And you can see it's fully draggable from from that perspective. Right?So I can move it here maybe and play around. And then I'm like, okay, let's make it a little bit more fancy. Now I have a GAN view as well. So the GAN view itself, it looks like this. You can actually click the task.You can play around with it. Everything is is right here. Now I said, great, but I really don't like the look and feel of it. So I asked Faze Studio, hey, can you actually be a UX expert and update the entire pages based on the current best practices? I want a whole new view.Right? I wanna remodel my house from that perspective. What Faze Studio was able to do is, first of all, give me a plan. This is going be a huge change for us. Right?So we need to have set of expectations. We generated a plan. What are we going to build? Why does it matter? What are the functional requirements that you we we can have here?What is gonna be the new user experience from for the user itself? What KPIs are we gonna impact from that aspect? Right? And then gave me the full architecture view, so everything would be aligned and recorded, and there is not not gonna be any any issues and problems later on. Everything was built.We approved the plan. And you can see here I made it I made it appear here very so you'll see it very quickly. But we effectively generated a whole new view for the app itself that looks like this. So all of a sudden the fonts are clean. It looks way better.It looks way nice. Right? The search and organization is different. Now I have a search bar that I can actually use here, which I did not have before. Right?I have a clear and visual view. I have a smart summary bar all of a sudden that I can sort by. You get a better feedback, like, you know, it was the cam and was lagging before. So now I have a better feedback option, and it it actually moves a little bit smoother from different tasks to different tasks. Right?Actually, do all tests so you can see it even better. And then guess what? That's like the big one. All of a sudden, it's actually mobile friendly. So we have this option to see everything on mobile and everything is gonna be fully supported for you on mobile.So that's kind of like the big impact that we can do. So instead of like hiring an engineer, they will understand what's going on the code base and do it back and forth and like, oh, I forgot to ask you this or I need to do that. I did everything here before before our call. And now everything is ready. And guess what?Instead of hiring a UX designer, I have a studio that understands the product itself and what you guys are doing. And I send everything to dev from that perspective. But wait, it's not over yet. So let's say I'm done with everything. I'm like, this is great.I love it. That's what I want to do. But you know, I want to start playing around with the fonts and do like fine tuning. All you need to do is just click design mode, choose an element on the page itself, and you can have a Figma style on your existing code. You can play around with everything.You can play around with the padding. You can play around with the colors. You can play around with the family. I don't know. Whatever you wanna build in here.I'm a very bad designer, by the way. And just one click of a button will change everything. Guess what? Maybe you wanna build a few variants of it. You have all the variants ready for you here, so you can play around and maybe play around with the flow itself.Everything in one place, All you need to do is just input everything everything in your native language, input a screenshot, input a Figma. I don't even care. You'll have everything on your code base, on your components, reusing your your engineering standards all the way to development.Jordan Metzner: Is this front end only? Is this connected to data sources? Kind of what is what is kind of the next steps of kind of turning this from kind of prototype into kind of the next level production? Or I mean, I saw the sent to devs, but, tell me a little bit more.Adir Ben-Yehuda: So it's it's a great question. It's not it's not front end only. It's front end and back end. You can actually everything is connected directly to the APIs. If you don't have APIs, we'll generate mock data for you and you can actually ask us to generate the the back end itself.But and that's a huge part of it. Right? Think about the audience we're catering. The audience we're catering is not like a audience that understand the code base itself. Right?So from a guardrail perspective, you won't do everything ad hoc. We wanna give a visual validation for the user itself. And once they decide to send it to dev, we have an agent that actually goes on the PR and connects everything together. You can be more prescribed and force us and ask from the agent to do everything. We have no problem in doing it as well.Jordan Metzner: Got it. So it can sit on top of your existing code base, creating new tools using the existing data sources you already have and then connecting it at the PR level?Adir Ben-Yehuda: Correct. You can update current screens or current flows from that perspective or generate a new one, but everything will be connected to your code base itself.Jordan Metzner: Awesome. Okay. Cool. And then deploy to dev is like giving basically taking that URL and putting it on the the Internet in some way or another. Right?Like giving you a host hosted version of that. Right?Adir Ben-Yehuda: Let's let's do it now. Let's send it to dev. It will take a few seconds. Right? And what it would actually do, it will prep the pull request itself.It will send it to your GitHub, your GitHub repository and will generate it a full pull request from that perspective. It it takes a few minutes because like we're preparing all the back end there from from the flow. And you will see it like every pull request you see from engineers. It goes to the CICD option. I'll give you a nice stat.On a weekly basis, we have a few thousands of pull requests. We have 86% approval on first shot. Because the pull request we're generating are part of your organization.Jordan Metzner: Oh, I got it. So, yeah, since the pull request knows kind of all the all the permutations that are required in order to have a successful PR, then it just does that the first time.Adir Ben-Yehuda: Correct.Jordan Metzner: And I've got one other question, Adira. I think we can kind of I think this has been an awesome demo. Tell us like, you know, what type of you know, what type of customers are using this? Are these like, you know, what type of enterprises are these? Like manufacturing?Have you found it better in tech? Is it more in, you know, industry or, you know, entertainment or kind of, you know, where have you found people have found, you know, kind of the best utilization for your tools? And maybe it's across the board.Adir Ben-Yehuda: Yeah. So today, as I mentioned, we have over a 160 clients. Most of them are tech native organizations, companies that are A, B, C around. Now we have a few enterprise organizations that have started using it, like big names that we're going to announce in the next couple of weeks probably. And in the last month, I think, started getting more approaches from, I would say, non tech native organization.Think about your banks or your financial organizations from that that option. So we're starting to see more and more people want to adopt it. Because there is one single thing we see across the board. People and companies are trying to enable the work upstream, right, the the digital people, the product people, the business unit or people inside the organization, and they don't have the right tools to do it. And that's where Faze Studio comes in.Sam Nadler: So I was, you know, this as we segue into kind of autonomy more as a whole as apart from this demo, you mentioned I think a 160 people. You you launched a 160 clients. You launched it about six months ago. What has surprised you in that six month period as as more and more customers start using it? You know, just getting their feedback, something that you didn't interpret they would use it.Were were new features launched based on their feedback, like that kind of the last feature about how you can get in and really, you know, change the design a la Figma. What are some of those things that evolved since launching?Adir Ben-Yehuda: I think the one thing that really surprised me is the how unhappy people are from Figma today. And it's it's crazy to see it. Right? Because you see the last report, they're skyrocketing in terms of revenue. But I think it's like just, I would say, not all the agreements that they had, but enterprise agreements that they still have in place.But if you look at the younger companies, they wanna get rid of Figma. They think Figma is not relevant anymore and what we saw across the board, and not what we've intended, is when companies are starting to use us, they're using less and less Figma from that perspective. I think that's the one thing. I think the other thing that surprised us, and not necessarily in the last in the last couple of months, but definitely in the last six months is the the, you know, the old guards, if you think about your Jira, even your linear organizations, people don't don't want to use it anymore. They just want to get they don't need to manage task in a more I mean, you know, cumbersome environment.They just need to act on the task, and that's what they need to do. Right? So when they started using us, they're like, oh, okay. We can open less task in Jira. We can manage that thing on Monday dev from that perspective.We can do everything directly with you, and that's about it. Right? So I think it was the impact of like, okay, everyone is talking about Cloud Code and everyone is talking about Codecs or Cursor from that matter. But that this like, I I think the amount of people that aren't happy from the old legacy products and the rate of innovation was just insane for me to see. Right?Like people are extremely not happy from what Figma is doing and and how innovative they are. People are extremely not happy from how linear and and are in terms of the rate of innovation, and they're just looking for every other solution under the sun.Sam Nadler: Very cool. So lastly, I wanna jump into the news quickly. But, you know, a, I think you touched upon it, but, like, tell us a little bit about the origin story of how like, what was the light bulb that went off that led you to, you know, launch this company?Adir Ben-Yehuda: Yeah. So me and my two other cofounders, right, I think the one thing we saw across the board is you people and companies are still giving more and more, I I would say, better shovels on the same workflows. And what we wanted to do is we wanted to challenge this. Right? I have I have a son.He's 10 years old. And he keeps building things on Unlovable and bunch of other things. Right? And then I ask myself, why can I have him work as a summer intern? Right?Like what kind of interface do I need to enable him? Right? This is definitely not child labor if this well, it is recorded. But you get the point. Right?It's it's like he can have fun. We can have fun together on those kind of things. But then the question we ask is like, why can we enable non technical individuals to really be part of the deployment? Why do we actually need engineers on on the flow itself? Why do we actually need all those cumbersome processes in the middle?And that's what we try to challenge. Right? And we started looking at the problem itself. We kind of like looked at the concept of really understanding the code base and how we can actually create real reuse of the code base itself. If you look at your current products, AI products, coding products, they're not really using the code base.They keep like running everything on the code base itself. Maybe they have a bigger context window, but they will rewrite components. They won't reuse components, right? And those are this is where problems are starting. When you have a person who is non technical, doesn't know what to do with it, and that's where where where it hits us.Right? We had a huge gap still in the market that we can come in and actually transform how our company works, not users, companies work.Sam Nadler: Amazing. Guys, okay. I moving on. I have two articles. Karpathy joins Anthropic.Is Karpathy known as the the creator of Vibe Coding, also one of the founders of OpenAI? Yeah. Obviously, a huge hire for Anthropic. What are your thoughts on what is he bringing to Anthropic? What does this mean for the space in general?Jordan Metzner: Yeah. I don't think he's a founder of OpenAI, but he was at Tesla. And I think Elon helped get him from Tesla over to OpenAI, and then, yeah, definitely defined the term kind of odd coding and obviously kind of many contributions in the space. But, yeah, I think it just just shows kind of how well Anthropic is doing in the market to be able to snag someone of his caliber. And obviously, we've seen the growth of Anthropic over the last few months in the enterprise space.So obviously, it just means that this is the beginning. You know, we saw, as I mentioned in the early in the episode, kind of the IPO coming for OpenAI, then I guess this definitely means kind of IPO probably coming for Anthropic as well. And, yeah, it still just feels like the early days, but I I think it's exciting.Sam Nadler: Didn't he hit the news just a couple months ago with some, I guess, research features? I forgot what what it was exactly.Jordan Metzner: Yeah. He, you know, he built this, like, tool to create your own, like, personal, like, notification like, note library that tracks what you're doing every day. So you have this, like, giant anthology of basically your entire life.Sam Nadler: It's gonna beAdir Ben-Yehuda: a huge market.Sam Nadler: Yeah. Yeah. Yeah. Cool. Guys, just to move on, and then, Dier, I'll start with you.I thought this was a a fun article. There was a ton of news out of Google in the last twenty four hours. You know, the new AI agents go beyond your standard searches. There's a lot to talk about with Google just recently, but, you know, I'm just thinking about, you know, flight searches and how you could theoretically, you know, immediately set up an agent in natural language. Hey.Let me know when flights to Tokyo get below $600 or something. You know, what do you guys think of the latest feature release across the Google Suite?Adir Ben-Yehuda: Yeah. I I I think from my perspective, we chose a clear trend. Right? Like people do not want to search anymore. They want answers.Right? They don't wanna see how the sausage is made. They want the sausage. And that's about it. I I think I mentioned this like a couple minutes ago about our product we filled, but that's the same thing.So I think it's like it's kind of a brutal irony that like perplexity in Chez GPT and every like AI startup spend the last two years being terrified from Google into building the like, you know, the Google product that they should have built a decade ago. And I think from that perspective, right now Google is trying to catch up with everyone. Right? So yes, they still have a huge distribution, but now their audience, they're starting to lose a lot of audience and they need to catch up from that flow from that flow and make sure that they give the audience what they want, which is effectively the results and not like booing or hyperlinks for that matter.Sam Nadler: Jordan, any comments on on this barrage of Google features?Jordan Metzner: Yeah. I'm long, Google. I think they'll figure out how to make it right. You know, at the end of the day, Google always wanted to answer the customer's question. I think it was just like how they answered the question.You know, they were always kind of unsure. That's why they gave you these, like, pages of links. But now it's easier to be more declarative on, like, the answer to things, and it seems like if you just want the answer, then you don't really want the other things around it. It probably will change how their ad model works significantly with probably, like, you know, higher intent in the sense that, like, you know, you may you maybe get less volume but higher intent. Right?So if you're looking for, like, mortgage broker near me, you know, you're probably gonna get a pretty high intent there. Yeah. I don't know. I think, like, you know, I've I'm about to test the new 3.5 flash model inside the anti gravity. I downloaded anti gravity.It's an entire new interface copying, you know, the the codecs and cloud code type interface. So I kind of lost the IDE part of it, and I don't know if I can get it back. But, yeah, I'm gonna play with it. You know, as you know, Sam, I'm you know, I've been favorable for the past few weeks slash months to Codex 5.5, but I'm definitely gonna test it out and see if it's good. Rumors are gonna I mean, Internet's kinda going back and forth on it.But, you know, obviously GoogleSam Nadler: Rumors more are better, right? Like, the latest rumors are it's better.Jordan Metzner: According to their according to their charts, but, you know, you gotta just test it and play in the game and see, you know, how it feels to you type of thing. But, you know, it's still early. They've got their TPUs. They obviously have the ability to, like, you know, bring token token costs down and, you know, and probably have high throughput. And so I think those two things could provide for a pretty dangerous recipe if they get a really good model on top of that.Yeah.Adir Ben-Yehuda: I think, Jordan, to your point about, like, the mortgage example. Right? If Google won't be able to deliver the mortgage itself, like not the middleman, not the middle person, right? Like actually say, hey, you want a mortgage, here's a mortgage from the best provider. All you need to do is like click next, next, next, sign everything, right?Connect your bank account and do that. That's where the world is going. That's the expectations from users that are now the 13 and 14 years old, right? And this is going to be the next buying power in the next decade. So I think that's exactly what they have to do.Otherwise, like from monetization perspective, you're missing the point of where the world is going.Sam Nadler: Alright, guys. Great episode. Adir, where can our viewers find you? Are you on x, LinkedIn, website? Let us know.Adir Ben-Yehuda: Yeah. So first of all, LinkedIn, feel free to connect. The dear Benny Houda, I will connect with everyone and like I'll just love talking and brainstorming with people. And of course, you can go to our website autonomyai.io and, you know, we have a free free trial. You can actually have a play around the environment that you can see the entire bells and whistles of the product and have fun.Sam Nadler: Amazing. Great episode, guys, and see you next week.Adir Ben-Yehuda: Thank you.Jordan Metzner: Awesome. Thanks, Adir. See you, Sam. Thanks, guys.Adir Ben-Yehuda: Thank you. Bye.
