Author: Nataraj Sindam

In the rapidly evolving landscape of venture capital, technology serves as the primary catalyst for innovation. Few understand this better than Gautham Buchi, the Chief Technology Officer at AngelList. With a rich background that includes senior roles at Coinbase and founding a Y Combinator-backed startup, Gautham brings a unique perspective on leveraging cutting-edge tools to solve complex financial problems. In this conversation with host Nataraj, Gautham dives deep into the operational core of AngelList, a platform dedicated to building the infrastructure for the startup economy.

He shares how AngelList is harnessing Generative AI to automate fund formation, provide deep, actionable insights for investors, and accelerate capital deployment. The discussion also explores the integration of crypto primitives, such as stablecoins and tokenization, to create new pathways for liquidity in private markets, a critical component for fueling the next wave of innovation. This episode is a masterclass in how modern technology is reshaping the world of startup investing.

→ Enjoy this conversation with Gautham Buchi, on Spotify or Apple.

→ Subscribe to ournewsletter and never miss an update.

Nataraj: To set the context of the conversation, can you give a quick introduction of who you are and what your journey was before joining AngelList as a CTO?

Gautham Buchi: My journey has largely revolved around key levers that can personally change someone’s life, which is largely education and access to financial tools. It started at Coursera, where we tried to democratize access to good education and then moved on to my own company, furthering the journey. Then to Coinbase, which democratized access to better financial tools using crypto as a methodology. Now I’m continuing on the path to democratize access to capital. Access to capital is probably the single best innovation hack we could do to create more startups. AngelList is in the business of creating more startups, creating more tools for the founders and builders. And I’m really excited to continue that journey there.

Nataraj: Talk a little bit about for those people who are not aware of AngelList. What are the different products on AngelList and what are the core business drivers among those products? You have rolling funds, venture funds, syndicates. Talk a little bit about that.

Gautham Buchi: A good mental model that I use is if you think of a triangle where one corner is the founders, the other corner is the GPs, like the Sequoias of the world, and the third corner is LPs, people who want to invest in early-stage venture or venture more broadly. AngelList is smack in the middle of the triangle. Our sole purpose is to make sure that the sides of this triangle are getting stronger and stronger because these three are the pillars of the innovation economy. The first thing you have to believe, to believe in the mission of AngelList, is that startups are good for the world. The creation of more startups is the way we innovate and is the way we accelerate innovation. Now post that, we need to identify how do we really strengthen each of these pillars: the founders, the fund managers, and people who want to invest in early-stage venture.

If you’re a founder, and maybe this is surprising to a lot of people, Robinhood’s first check was on AngelList. Many companies through our product have been able to come in and say, ‘Okay, I’m looking to access good capital, not just dumb money, but good capital on the platform.’ I can go to AngelList and start a company. We are creating an ecosystem for founders to take the mental gymnastics around starting a company and really focus on building the product. We will build the rails for you to get the capital that you need.

Moving on to the other corner of the triangle, which is you are a fund manager. Let’s say you have a unique hypothesis, a unique insight into where you think you can be investing to accelerate this innovation. You need two pieces: access to good founder opportunities and access to good capital that is looking to be invested. That is our core fund admin product, the core GP product. This is probably the one that AngelList is well-known for today. You get a lot of tools so you don’t spend time doing the gymnastics of how to raise and deploy capital, but really focus on what you can do to add maximum value to the founders.

The third corner of that pillar is the people looking to invest in early-stage venture more broadly. This is probably the one that most people have historically known AngelList for. Wherever you’re in the world, if you want to invest, get access, and believe in the startup economy, you can write a $1,000 check to a $100,000 check. You want to be an angel, you go to AngelList. This is the thing that Naval envisioned: how do I really democratize access to early-stage venture across the world? So we provide a number of tools for people who are looking to get their toes wet in the world of angel investing. To sum it up, the way I would think about AngelList as a business is to really think about the triangle between founders, GPs who are looking to run the fund, and then angels. The speed with which we can spin the triangle is essentially innovation.

Nataraj: You joined AngelList this year or last year, and you’ve worked in different companies. How is working at AngelList different from working at Coinbase?

Gautham Buchi: Very different. Right off the bat, crypto in 2017-2018 was very different than crypto right now. To give a specific anecdote, if you and I met in 2017 and you told me that by 2025 we would have a Bitcoin ETF or we would have stablecoins, most people in crypto would have laughed. The pace of innovation is so constant, so relentless, and quite frankly, very uplifting. But there is always this overhang of regulation on top of you. There were many times, especially over the last four years, where being at Coinbase felt very much like you’re fighting a big institution, a regulatory battle. That is not something that we face at AngelList. You don’t spend time thinking about regulation in the way that you would in the crypto world. You’re really thinking about how do I accelerate capital deployment? How do I bring more efficiency to how capital is being deployed? Which is a very different problem space.

Second, the pace of innovation in crypto is insane. We had a joke at Coinbase that one year in crypto is like 10 years. There’s a popular meme where after five years in crypto, somebody has this white beard and gray hair. It’s very true. I can personally attest to it. And so is the eternal optimism. The crypto crowd is probably one of the most optimistic crowds that I’ve ever worked with. It’s different in capital products. While innovation does happen, it’s not at the same pace at which it’s happening in crypto. So the way you think about product, you’re thinking more from a reliability lens, you’re thinking longer-term, which is very different. As for the companies, particularly, AngelList is much smaller, much more early stage. We are about 150 people. Coinbase, when I joined, was probably a few hundred, but it’s now a much bigger company. So that definitely has its own pros and cons.

Nataraj: There’s one through-line I see between Coinbase and AngelList: both were involved in major regulatory changes. Naval and the team were involved in the JOBS Act earlier to change and make AngelList and crowdfunding happen. Now we are seeing that happen in real-time with some of the crypto legislative changes. I want to pivot towards what I wanted to talk about most in this conversation, which is about AI. Post-ChatGPT, we saw you could do a lot more with this current technology. In my career, this feels like a game-changing moment. I wanted to quickly get your thoughts on what you think of Generative AI and this current AI hype cycle.

Gautham Buchi: Let me dial back the clock a little bit. I don’t know if your audience is familiar with Coursera; it was an education platform that started in 2011. Our first major success was a machine learning course by Andrew Ng. A lot of people, especially in deep learning, probably got their start with Andrew. At Coursera, we were incredibly excited about it, not just from a pure technology perspective, but also the audience and the learning; these were the most subscribed courses on the platform. The thing that was different back then was it was still largely a research problem. It was harder to think about what an actual go-to-market version was.

Even when I was starting my own company in 2016-2017, there was a running joke within YC that all you had to do was attach .ai to your domain and you automatically would raise a bunch of money. So there was definitely hype cycle number two happening in 2017. What’s different about this particular iteration is one, it moved from being a research problem to an engineering problem. You could take the model in a box, assume somebody has already done all the heavy lifting, and now you’re just trying to figure out what other things in the ecosystem you need to connect to make sense out of this. That’s been incredible to see.

Second is the utility of it. The utility back in the first cycle of my experience, 2012-2013 machine learning, 2016-17 AI, you really had to squint your eyes. There was always a human in the loop. The utility was not obvious. You had to bet that one day this thing would actually be at a point where you will see real feedback loops. But we are in a world where you can parse a PDF instantly in a couple of seconds, or you could do voice translation. So now you bring these two ideas together: it’s an engineering problem, and the utility is instant. That means you have a very fast feedback loop. You and I can spend the next 20 minutes and literally build something, put it out in the world, and see how people are interacting with it. And that is very powerful.

Nataraj: What do you think of the use cases that are most exciting for you as a CTO, and how are you at AngelList adopting AI in different ways?

Gautham Buchi: There are two interesting questions there: what is very exciting to me, and what is very exciting to the business. To me, it is so interesting to see the blurring of the roles. Even three years ago, if you wanted to build an MVP, you’d ask, ‘Who’s going to be the designer? Who’s writing the PRD? Who’s going to be building this project?’ That’s a lot of overhead. Today, our chief legal officer builds an end-to-end product himself. That includes the design, the spec, the PRD; he releases it, he’s tracking analytics. Our designer is building end-to-end products. An intern is building end-to-end. We really went from a role in a box to a product in a box. We have this full spectrum of skills that are very much available to you. The conversations become so much sharper on a day-to-day basis. This idea that you have to go through multiple iterations to even define what you’re doing will become so outdated, and the roles become so blurry. It is increasingly becoming hard to define the role of a product person versus an engineer.

Second is your ability to deploy and get the boilerplate out of the way has been huge. The hard problem in most companies is working with legacy code, not greenfield code. The moment you are able to put things in place that can abstract the legacy away from you or even better, intelligently retool the legacy for you, you’re taking a ton of work out of the way. We are able to now see folks join and start deploying the same day. It used to be aspirational, but now it’s almost an expectation because of all the tooling available.

The third thing goes back to the base-level expectation. I have this view that it will be increasingly hard to see a good role for yourself if you don’t become very quickly AI-native, meaning being able to understand which tools create maximum leverage. It almost feels like, ‘Am I late?’ You’re not late, but now is a good time to start. I can clearly see the difference between teams that have adopted AI and the teams that are still lagging behind. The difference is so clear, so obvious, that we now have a default expectation that everybody’s trying out these tools.

Nataraj: What are the blockers for teams that aren’t AI-native?

Gautham Buchi: I don’t think it is a philosophical stance. It is more of an inertia and momentum thing. You could also be skeptical. For what it’s worth, I was skeptical at one point as well. If you are an engineer today and you have not tried one of the IDEs like Cursor, CodeWhisperer, or Copilot, you are already behind. So inertia could be a big component. Second, there are some good reasons not to do it, depending on which team you’re in. For example, if you’re in security or a very critical path, you want to spend that extra time and attention. At Coinbase, we had a lot of concern internally around what we might accidentally expose because a lot of these are also primitives that are being built right now.

Nataraj: I always call AI right now ‘draft AI’ because it gets you the draft pretty fast. But if I’m reporting business numbers to my leadership, I want to depend on myself to review each line, even if I use AI to write it. You still need that 5% manual intervention, but that 95% is a really big time saver. Can you talk about some examples of how you’re using AI in your own products at AngelList?

Gautham Buchi: Let’s talk about our customer type. On a typical fund deployment, there are a lot of workflows you go through that are sequential, whether it’s legal, boilerplate, or dependent on internal movements. One of the metrics we track religiously is how long it takes for you to deploy your fund, raise your fund, or get set up for the fund. We are increasingly using a lot of AI and automation to do that. One thing we do is doc parsing. In a fund formation, there are tens, if not hundreds, of docs. We can parse the docs, provide the information that is very relevant to you, and automate your deployment. This is integral to how we simplify fund formation.

The second thing is operational. Once you have your fund deployed, the thing that AngelList is known for is the venture associates and the quality of service. We want to enable our internal teams to very quickly get access to data at their fingertips. A couple of years ago, pulling up a specific legal term for a GP would be half a day’s task. Now we have built internal tooling where our customer support and venture associate teams can, in most cases, auto-resolve issues. We can pull information, make sense of it, and spit out exactly what the customer is looking for. This feeds into the cycle of closing feedback loops and becoming more efficient.

The third bucket is AngelList is sitting on a gold mine of data. Some of the hardest resources to get to is early-stage venture data. There are hundreds and thousands of companies on the platform raising money. There is a tremendous opportunity here where we can drive deep insights into what’s happening with your portfolio. We can tell you exactly where you’re invested, opportunities you might be missing, and how your fund is performing compared to the rest of the funds on the platform. We are now able to start doing some of that using AI.

Nataraj: Talk a little bit about your crypto integration as well. I know AngelList was one of the first adopters of Circle a couple of years back. Is tokenizing shares on a blockchain a path that AngelList is looking towards?

Gautham Buchi: This is one of the best opportunities for AngelList. One thing we have done very concretely today is we enable USDC funding. If you’re a startup that is raising money with USDC, AngelList allows you to do that at no fee, and we have seen pretty significant adoption. The second opportunity is distributions. For a lot of crypto companies, distributions happen through crypto tokens. Us being able to support that means if you’re a crypto company that has an exit, your investors are able to get and keep those tokens on the AngelList platform.

Moving on to stablecoins, I think it’s one of the most exciting areas right now because they’re instantaneous, near-instant settlement. This drastically simplifies cross-border wires, which is a massive pain. This is something we are seriously thinking about: how do we make capital deployment more efficient? We are seriously thinking about how we can make stablecoins a primary citizen on the platform, potentially enabling digital wallets for all customers and LPs.

The second bucket is tokenization. What has changed over the last seven or eight years is companies are increasingly choosing to stay private. Stripe, OpenAI, Anthropic are examples. This means your capital is locked for much longer than historically seen. While you’re happy the valuation is going bonkers, at the end of the day, this is on paper; it’s not liquidity yet. And liquidity is really important because it fuels the next generation of startups. One of the things we are seriously thinking about is how do we create liquidity for the GPs and investors on the platform. On the technology side, tokenization is a reality. We are seriously thinking about how we can bring the regulatory framework, tokenization framework, and KYC/AML together to create liquidity for the funds on the platform and create good incentives for founders to participate in it.

Nataraj: As the lines are blurring, what skills should product managers invest in building?

Gautham Buchi: The thing that has changed is your ability to go from an idea to seeing it in the world has dramatically changed. The most powerful thing product managers have today that they didn’t have before is an ability to take their product idea, put it out in the world, gather actual data, and then come back to the table. They can say, ‘Here’s an MVP that I was able to build for myself. I’m not stuck in multiple rounds of prioritization. Here are 10 people I have shown this to, and here’s the information I received.’ That is so powerful and empowering. The classic role of a product manager as an information router is quickly disappearing. If you are purely serving the purpose of routing information and doing prioritization, you’re in trouble. We have moved to a world where product managers are empowered to very quickly generate these prototypes and take them to market. That’s what I would invest in right now.

Nataraj: Thanks, Gautam, for coming on the show and sharing your insights and time.

Gautham Buchi: Likewise, thank you for having me and nice to meet you all.

Gautham Buchi provided a clear look into how AngelList is pioneering the future of venture capital by integrating AI and crypto. This conversation highlights the tangible benefits of these technologies in making startup investing more efficient, accessible, and liquid for founders, GPs, and LPs alike.

→ If you enjoyed this conversation with Gautham Buchi, listen to the full episode here on Spotify or Apple.

→ Subscribe to ourNewsletter and never miss an update.

The world of digital advertising is a complex, rapidly evolving ecosystem that powers much of the free content we consume online. At the heart of this system is programmatic advertising, a technology that automates the buying and selling of ad space in real time. In this conversation, Nataraj sits down with Jared Siegal, the founder and CEO of Attitude, a company at the forefront of empowering publishers to maximize their revenue in this competitive landscape. Jared shares his unexpected entry into the ad tech industry, demystifies concepts like header bidding and ad exchanges, and explains how his company’s SaaS model is disrupting the status quo. They also explore the seismic shifts caused by AI in search, the ongoing debate around Google’s market dominance, and what the future holds for content creators and publishers trying to navigate this intricate digital world.

→ Enjoy this conversation with Jared Siegal, on Spotify or Apple.
→ Subscribe to our newsletter and never miss an update.

Nataraj: So I think a good place to start would be how did you get into this ad business?

Jared Siegal: Totally by accident. I don’t think anyone grows up saying, “I’m going to serve ads on the internet” or even really understands that this part of the economy exists. I went to school for econometrics, which is applying economic theory to math problems and vice versa. I couldn’t even tell you how I got into that, but as I was getting ready to graduate, I really wanted to work in the car industry. I reached out to every graduate from my university who worked at Ford, Chevy, and all the major brands here in the US. Couldn’t get a job.

I went to the head of our school’s entrepreneurship program and said, “Hey, I got a cool idea for a class I wanna teach here.” I pitched it to him, and he said, “This is a great idea, but I really think you should meet this guy, a former graduate from our school. He runs a company called Answers.com.” I met him, and frankly, I had no idea what they did. I didn’t understand it. He offered me a job and I said sure. And that’s how I got into online advertising. I had a choice of working on the revenue side of the business or the cost side. I was always taught growing up to always be a revenue driver, so I chose the revenue side. That forced me into Ad Ops, and very quickly, within a few months, I fell in love with this industry.

Nataraj: So what were you doing? Was it trying to grow revenue or grow traffic?

Jared Siegal: It was twofold. One was actually trying to grow revenue, and one was trying to grow traffic, which obviously indirectly and directly grows your revenue. On the revenue side, this was right when DFP, now GAM, was created. So I was literally learning how to integrate DFP on a website, figuring out how to get away from this concept of a waterfall auction into something a bit more programmatic and real-time, and creating a bunch of different layouts and page types to understand which ad units, sizes, and arrangements make us the most money.

Nataraj: Can you explain DFP and the waterfall concept?

Jared Siegal: Yeah, DFP, which is now called GAM, is Google’s ad server. It’s used by almost every website on the planet to host the final auction of that ad on your website. Before that, people were hard-coding ads on their websites and hoping they made money. The creation of the ad server meant that you as a publisher could host an auction, get a bunch of people to compete for that ad, and choose the highest winner. Waterfall is this idea of, let me call Google, if Google doesn’t fill, let me call partner B, if partner B doesn’t fill, let me call partner C. Where we are today in programmatic is, let’s get Google, partner A, B, C, D, E, F, G, all to compete in real-time. They all bid at the same time, and whoever wins, wins. It’s a little bit faster, a little bit more efficient, and it’s far more accurate in terms of valuing your audience.

Nataraj: So let’s explain the lay of the land today. For example, I go to a site like verge.com and I see display ads. What are all the players involved when I’m seeing that ad? Who’s the publisher, who’s the bidder, who’s the exchange, and what is Google’s role versus Attitude’s role?

Jared Siegal: Okay, cool. So you go to that website; the website is the publisher. They’re the one that is publishing the content, and you’re on that site because you like their content. When that ad gets served, 99% of the time that publisher probably uses Google Ad Manager as their ad server. It’s what eventually makes the final decision of who had the highest bid from all of these exchanges. Google is also an exchange, but there are hundreds of exchanges that work with publishers directly and tens of thousands behind the scenes. All of these exchanges need what’s called a wrapper to host this auction and pass all the different bids and ad creatives into Google Ad Manager so it can make a decision. And that’s what Attitude does. There’s a handful of companies that do that part of the business. You have the ad server, you have the advertisers, and you have the company that is connecting the advertisers to the publishers. Attitude is that connection.

Nataraj: So you collect the different bids for the ad spot. Where are you collecting them from?

Jared Siegal: It’s all happening in the browser in real-time. Publishers basically load our code in the head of their page. On page load, boom, we instantly start pinging all these different advertisers they have relationships with to find the highest bids and send them along. It’s happening in milliseconds. For that one ad to be served to you on that one website, there were probably millions of different agencies, brands, and companies that got pinged in a matter of milliseconds to say, “Do you have something for me?”

Nataraj: Why do customers choose Attitude versus just using Google? Because it feels like Google is a competitor here.

Jared Siegal: To some extent. You could go directly with one exchange, and they’ll probably be able to serve most of your ads. But what happens when you only have one exchange is it’s no longer really an auction. They can pay whatever they want for that ad because they don’t have to beat out anyone else. Where a company like Attitude comes into play is we say, don’t let Google or Facebook dictate the value and price of that ad. Have a bunch of people compete and let the highest one win. In an auction, you want as many bidders as possible. You don’t want one person bidding because then they’ll just bid a dollar and they’ve won.

Nataraj: So it’s better to use Attitude to create a neutral playing field.

Jared Siegal: Yeah, you need some piece of technology to do that because Google Ad Manager and most ad servers don’t natively integrate all of the other exchanges. They’re limited to their own exchange. So if you want a bunch of exchanges to compete, you need this third-party tech to layer on your page. How we separate ourselves is our business model and the fact that we are agnostic. Everyone else in our space takes a percentage cut of the publisher’s business. We don’t have ulterior incentives to let one exchange win more because they pay us a higher rev share. It’s irrelevant to us who wins. We just want the publisher to make as much money as possible. We built a pretty big name for ourselves as the first SaaS pricing model in this space.

Nataraj: Let’s talk about Google’s role. Do you have a view on the whole trial of Google as a monopoly?

Jared Siegal: Let me preface this by saying we’re a really good partner of Google’s, and Google’s a great partner of Attitude. But there’s a reason why companies like mine exist, and that is because Google has historically had the last look at every auction. If they’re the ad server being used, they see all the other bids that come in, and after they see all that, they can say, “Hey, do I want to bid one penny more and steal that impression?” That starts getting into this idea of, is it really a fair auction? Companies like mine have been coming up with creative ways to make it fairer, whether it’s through setting price floors or creating our own ad server. With all of the recent news about monopolization, if you’re in our space, you’re kind of sitting back saying, “Yeah, obviously this has been going on for 20 years. Everyone knows this.”

Nataraj: How did you start as a consulting firm and transition to a full-fledged product company?

Jared Siegal: I started this company by accident. I quit my job and just wanted to do something on my own, so I started consulting for a bunch of publishers I had become friendly with, charging them by the hour. I did that for about 12 to 14 months and got the business up to close to a million-dollar run rate. Back then, auctions were second-price, meaning the winner pays one penny higher than the second-highest bid. I made a career for a year of trying to figure out the gap between the first and second bids and setting minimum prices to capture more revenue. Then Google said everything’s moving to a first-price auction, and my whole business model was gone. At that time, a lot of my clients were using the same header bidding company and having a lot of issues. They were paying me an hourly rate to communicate those issues to this third-party company. I realized, why am I helping someone else grow their business? I should build this piece of tech myself, do a better job, and sell it to my existing clients. I gave it away for free for six or seven months to grow the tech, and eventually, I converted all my clients. At that point, I got an offer to buy the company from an ad exchange. I was blown away. I sat down with my wife and some friends, and they all said, “Don’t sell, grow the business.” So I called up my best friend, who’s now our CTO, and said, “Quit your job. Come over here. Let’s build something.” And the rest is history.

Nataraj: Post-ChatGPT and Google’s AI search results, how is that affecting publishers?

Jared Siegal: For sure. The fact that Google rolled out AI in its search results radically changed SEO. If you’re a website where the majority of your content is easily answerable in one sentence or a yes/no manner, AI is going to crush your business because the answer appears in the search results and the user never clicks through to your website. If you’re a site that has opinionated, long-form content, or things that are not a simple question-answer relationship but more like thought pieces, you’re probably much safer, at least for now. AI inside of search results has made the internet worse. I think most publishers would agree. Every piece of tech developed in our industry has always been to help the biggest players—advertisers and search engines—not publishers. AI has a huge impact on traffic for a lot of publishers.

Nataraj: Internet traffic seems to be shrinking or consolidating, but Google and Facebook are still increasing ad revenue. How is that possible?

Jared Siegal: To some extent, it’s pricing control, but also an important piece of information is that any search engine probably makes more money from the ads served in their search results than the revenue share they get on ads they help serve on publisher websites. If you search on Bing and click on one of the paid search results, they probably made a dollar. If you click a link to a publisher’s website, they might make a few pennies. There’s a huge asymmetry and a conflict of interest here. It behooves them to not send you traffic and to keep you within the search results page. They make more money that way.

Nataraj: What’s a common misconception about running a company that you’ve found not to be true?

Jared Siegal: There’s this concept that was hot a few months ago about founder-led versus employee-led businesses, and many people were anti-founder-led. I am very involved in the day-to-day of Attitude, from cutting checks to talking with publishers to running A/B tests to negotiating deals. I love it and I do it all. I think a successful entrepreneur and leader is someone that actually understands all aspects of the business. People say, “Just hire smarter people and have them handle all that.” 100%, have them handle it, but you better understand what they do better than they do. If you want to run a successful company, you need to understand every penny that comes in and every penny that goes out. We’re very much a founder-led business, and I think it’s what has allowed us to scale up as quickly as we did.

Jared’s insights reveal the intricate balance of power in the ad tech industry and the critical need for solutions that champion publishers. As AI continues to reshape content discovery, the strategies discussed in this conversation offer a valuable roadmap for navigating the future of digital monetization.

→ If you enjoyed this conversation with Jared Siegal, listen to the full episode here on Spotify or Apple.
→ Subscribe to our Newsletter and never miss an update.

The global waste crisis is a staggering $1.6 trillion problem, with mountains of discarded materials ending up in landfills and oceans. But what if we could see this “waste” not as trash, but as a valuable resource? This is the mission of Ambarish Mitra, co-founder and CEO of Grey Parrot. After a successful journey in augmented reality with his previous company, Blippar, Ambarish pivoted to tackle a more tangible and pressing global issue. Grey Parrot uses sophisticated AI and computer vision to analyze and sort waste streams in real-time, bringing unprecedented intelligence to the recycling industry. In this conversation, Ambarish discusses the technological challenges of deploying AI in harsh industrial environments, the importance of building a cost-effective hardware and software solution, and how data is key to unlocking a truly circular economy where materials are recovered and reused, not discarded. It’s a fascinating look at the intersection of deep tech and environmental sustainability.

→ Enjoy this conversation with Ambarish Mitra, on Spotify, Apple, or YouTube.
→ Subscribe to ournewsletter and never miss an update.

Nataraj: What is Grey Parrot, and how did the idea start?

Ambarish Mitra: Grey Parrot is a waste intelligence platform that uses computer vision-based AI blended with material sciences to recognize large-scale waste flows. When people throw away rubbish, it ends up in material recovery facilities where it’s processed and sorted for recycling, landfill, or incineration. Grey Parrot uses analyzer boxes to recognize 100% of the waste flowing through these plants, helping to sort it more efficiently. It’s a large and complex problem because humans generate garbage at such a massive scale that it can’t be solved with just human or mechanical interaction alone. It requires a large amount of vision-based processing and was almost waiting for the AI era to kick in to address it. We saw a large, unaddressed opportunity. Plus, waste is a global crisis that impacts lives and the planet, so we decided to address this issue head-on.

Nataraj: Was the initial idea to do what you’re doing today, or was it different?

Ambarish Mitra: It was different. My co-founder and our initial team came from my previous company, Blippar, where our mission was to build the world’s first visual search engine. We built a large-scale vision model, but we realized our revenue model led to recognizing brands that often ended up in the bin. This got us thinking. Everyone has mapped the consumption world—Amazon, DoorDash, Instagram all know what you’re about to purchase. But after that $23 trillion of annual consumption ends up in the bin, there was almost no digitization. I call it the shadow economy. One reason waste remains waste is that no one is doing enough digitally to value and recover it. That’s why so much value is lost. So the idea came: why don’t we use our vision expertise to do something more impactful and circular? We call it waste, but we see it as paper, aluminum, and different types of plastic. We think of ourselves as a material asset recovery company rather than a waste company.

Nataraj: What is the actual product that you’re selling to companies in the recycling ecosystem?

Ambarish Mitra: Let me give you a brief intro to how waste works. Waste is thrown in bins, collected by trucks, and taken to Material Recovery Facilities (MRFs). It’s tipped out, piled onto conveyor belts, and goes through layers of mechanical processes. There are large leakages in that process, and a majority of that leakage ends up in landfill. Our goal is to reduce that leakage. We built hardware we call the analyzer. The job of the Grey Parrot analyzer is to analyze 100% of the waste flow in real time. These are rivers of waste on belts two meters wide, moving at three meters a second, processing up to 1,500 tons of waste per day.

When the camera recognizes 100% of the waste flow, it helps plant owners understand the unit economics of their business—what material comes through and what its financial value is. Secondly, it provides waste analytics to show if the plant is efficient or inefficient because every percentage difference is a revenue opportunity. The last thing is quality control—the purity of the materials. The more single-stream a material becomes, the more a buyer will pay for it. Finally, we’re integrating a brain into these mechanical machines, much like Waymo makes existing cars into self-driving cars. We are making these plants semi-automated by applying intelligence to existing mechanics, sending signals from one gate to another to ensure everything is sorted as purely as possible. The plant owner sees a dashboard where all this data is available, showing if the plant is working optimally.

Nataraj: What are the architectural and structural issues specific to this industry that you had to navigate? It sounds like you’re shipping hardware and software into environments that are not known for being tech-savvy.

Ambarish Mitra: That’s a great question. This is not a category where you can grow at any cost. It’s a cost-prohibitive industry where every cent matters. Unlike growth-oriented industries like e-commerce or advertising, you can’t have a variable cost architecture where revenue compensates for growth costs. Here, we have to recover more waste and create value from it. The tonnages are massive. So, we had to build an architecture where a lot happens locally on the machine. Our deep learning models sit locally so our costs don’t go up as we process hundreds of millions of images. We process images at the scale of social networks, but we’re processing trash, not people.

It also needs to be near real-time, because the system has to react within 30 milliseconds to trigger a robotic arm, an optical sorter, or stop the plant for hazardous materials. The system cannot rely solely on internet connectivity. We came up with an architecture that requires the internet periodically, but a lot of the processing is on the edge. A huge amount of the vision processing actually happens on the camera itself to normalize images, because lighting conditions in every plant are different. We built one platform that works in every plant. It was an interesting challenge to consider everything from image capture to model building to ensure it works with 99% efficiency, 24/7.

Nataraj: Can you talk a little bit about customer acquisition? How did you approach your first five to 10 customers and how do you scale now?

Ambarish Mitra: As an outsider, we had to learn the hard way. We came from a background of large-scale, vision-based compute, but we didn’t understand waste. So, in the first days, we did something smart: we built the first version of the product *with* the waste industry. We asked waste management companies what problems they were trying to solve, like counting for audit trails or quality control. We learned from them and released our first version by talking to seven or eight customers, giving them the intelligence for free for the first two years while we built our larger model.

We also didn’t build it in just one geography. We spread out across Europe, America, and South Korea to get diversity of data. Commercially, we started with a direct sales model, hiring people from the industry. Then we learned there’s a whole middle tier of specialized salespeople who are plant builders. They were already aggregating multiple technologies to build a plant, so it made sense to partner with them. In the last two years, we partnered with Bolograph, the world’s biggest plant builder, and Van Dyke Recycling Solutions in the US, America’s largest. We disintermediated our direct sales model through these strategic partnerships, which made us more cost-efficient and allowed us to scale effectively.

Nataraj: Which countries are doing the best when it comes to waste management?

Ambarish Mitra: Japan and Korea are very good. Germany is very good. The society is very conscious, and it’s designed to collect waste in many forms, not just from bins. Germany has a direct deposit scheme where people can return bottles for vouchers, for example. I would say there are four components to solving this. One is the manufacturer, who can take more responsibility through standardization, like how USB cables were standardized. Then you have the government’s role, which can enforce regulations. Then you have the waste management side, which can optimize and digitize with AI. And the last quadrant, which has a lot of power but often doesn’t use it, is the consumer making choices that are more circular in nature. Today, consumers are making some choices, governments are doing something, and a few brands are doing a few things in fragments, but a perfect storm hasn’t happened yet.

This conversation with Ambarish Mitra offers a compelling look at how advanced AI can be applied to solve one of the world’s most fundamental environmental problems. Grey Parrot’s innovative approach not only enhances the efficiency of recycling but also provides the critical data needed to build a sustainable, circular economy for future generations.

→ If you enjoyed this conversation with Ambarish Mitra, listen to the full episode here on Spotify, Apple, or YouTube.
→ Subscribe to ourNewsletter and never miss an update.