Snapchat SVP of Engineering on How a Billion-User App Lets Everyone Ship Code Without Breaking Quality | Saral Jain | E304

In this episode of The Product Podcast by Product School, Carlos González de Villaumbrosia sits down with Saral Jain, SVP of Engineering at Snapchat, the last independent social platform operating at global scale, with 956 million monthly active users closing in on the one billion mark and a community that opens the app more than 30 times a day.

Saral joined Snap nine years ago, right around the IPO, after nearly a decade leading engineering teams at Amazon Web Services, and today leads all engineering for Snapchat, spanning product experiences, multi-cloud infrastructure, and the company's machine learning and generative AI platforms.

What you'll learn:

• How Snap lets designers and product managers ship production code, with an AI agent running the first review pass on 90% of code within 5 minutes
• What Casper is: the AI teammate any team at Snap can invoke from Slack or Jira to build a working prototype from a conversation
• How Snap turned company-wide AI adoption into business impact after early prototypes were being built and thrown away
• How lean startup squads mix engineers, designers, and data scientists to launch zero-to-one bets inside a mature platform

Key takeaways:

• Quality control does not have to slow down who gets to ship, it has to change what reviews the work first
• Widespread AI adoption is not the same as business impact, and the gap between the two is where most AI investment is being wasted right now
• Small, cross-functional teams with blurred roles can move faster than traditional org structures, even inside a billion-user company

Credits:
Host: Carlos Gonzalez de Villaumbrosia
Guest: Saral Jain

Social Links:

• Find out more about Product School here
• Follow our Podcast on TikTok here
• Follow Product School on LinkedIn here

Transcript

Introduction

Saral Jain | Snapchat 00:00:00 The bottleneck is not Figma mockups. The bottleneck is we just build the prototype, and then show the prototype, and the best idea wins. It reviews more than 90% of the code within five minutes of the code being published. Why don't you go ahead and build it? And then Casper goes ahead and builds it. It's a super wild time to be an engineer.

Saral Jain | Snapchat 00:00:17 Adoption wasn't the problem. Translating that adoption to business impact was the problem. The boundaries between these roles will continue to become less and less clear, and I don't think it matters what your title is. It really matters what you're spending your time doing.

Carlos González de Villaumbrosia | Product School 00:00:34 Hey, this is Carlos, CEO at Product School and your host on the Product Podcast.

Carlos González de Villaumbrosia | Product School 00:00:38 Today's guest is Saral Jain, Senior Vice President of Engineering at Snap. Snapchat is the last independent social platform operating at global scale, with 956 million monthly active users closing in on the one billion mark, and a community that opens the app more than 30 times a day. The company generated close to six billion dollars in revenue in 2025 and grew 12% year over year last quarter while returning to daily user growth.

Carlos González de Villaumbrosia | Product School 00:01:03 Saral joined Snap nine years ago, right around the IPO, after nearly a decade leading engineering teams at Amazon Web Services. Today, he leads all engineering for Snapchat, spanning product experiences, multi-cloud infrastructure, and the company's machine learning and generative AI platforms. In our conversation, we cover letting designers and product managers ship production code, with an AI agent running the first review pass on 90% of code within five minutes.

Carlos González de Villaumbrosia | Product School 00:01:29 Casper, the AI teammate that any team at Snap can invoke from Slack or Jira to build a working prototype from a conversation. Turning company-wide AI adoption into business impact after early prototypes were being built and thrown away. Running lean startup squads that mix engineers, designers, and data scientists to launch zero to one bets inside a mature platform.

Carlos González de Villaumbrosia | Product School 00:01:49 Let's get into it. Welcome to the Product Podcast, Saral.

Saral Jain | Snapchat 00:01:54 Thanks for having me. It should be fun.

A Billion Snapchatters and Fifteen Years of Staying Independent

Carlos González de Villaumbrosia | Product School 00:01:56 It's been fun to dig a little deeper into Snapchat and, first of all, realize that the company's almost 15 years old.

Saral Jain | Snapchat 00:02:05 That's true. Most people don't realize we've been around for a while.

Saral Jain | Snapchat 00:02:08 It's almost 15 years old, and we have thrived to almost a billion Snapchatters in our community right now. So it's been a fun ride.

Carlos González de Villaumbrosia | Product School 00:02:16 You said over a billion users?

Saral Jain | Snapchat 00:02:20 Almost a billion users. So we have around 953 million monthly active users right now, using our platform in a very, very engaged way.

Saral Jain | Snapchat 00:02:27 So almost a billion people use Snapchat.

Carlos González de Villaumbrosia | Product School 00:02:30 That's crazy, and that makes me feel a little old because I was an early user, but I have to confess.

Saral Jain | Snapchat 00:02:35 You and me are both in the same boat there.

Carlos González de Villaumbrosia | Product School 00:02:38 And so you've also been at the company for, what, nine years?

Saral Jain | Snapchat 00:02:42 That is correct. I joined right around the IPO, so I've been here around nine years.

Saral Jain | Snapchat 00:02:46 I played various different roles at the company as well.

Carlos González de Villaumbrosia | Product School 00:02:49 And now you're currently SVP of engineering, right?

Saral Jain | Snapchat 00:02:52 That is correct. I lead most of engineering for Snapchat.

Carlos González de Villaumbrosia | Product School 00:02:55 Help me understand the org design. So I understand you report directly to the CEO, and then how is your team structured?

Saral Jain | Snapchat 00:03:03 I think we have a very functional organizational structure, where all of the functional leads, whether it's the head of product, whether it's engineering, whether it's sales, legal, marketing, or comms, all of those functional leaders report to our CEO. So we don't have the typical general manager model, where there's a single accountable owner.

Saral Jain | Snapchat 00:03:21 For all major initiatives, we have to work across all of our teams to get stuff done, which has pros and cons, but it really works very well for a company like Snapchat to have functional leadership report to our CEO.

Carlos González de Villaumbrosia | Product School 00:03:33 He's effectively the CPO as well, I guess?

Saral Jain | Snapchat 00:03:37 He is. He's a designer by heart, by training, and he's our chief product officer and design officer as well.

Saral Jain | Snapchat 00:03:42 Although we do have a head of design and a head of product that report to him.

Engineering, Product, and Design as Peers

Carlos González de Villaumbrosia | Product School 00:03:45 So let's talk about it. What's the relationship between product management specifically and you as the engineering leader?

Saral Jain | Snapchat 00:03:52 It's a very, very tight partnership. Almost everything we do, we do in very, very close coordination.

Saral Jain | Snapchat 00:03:58 It's a very design-focused company, and so a lot of the time the ideas come from our design team and the product team, who really pay close attention to our community and the pain points of the community and what we want to create for our community. And so we have very clear long-term thinking about what we're trying to build.

Saral Jain | Snapchat 00:04:15 So the product team comes up with some of the ideas, the design team comes up with ideas, and the engineering team also owns some of the ideas, and we all work very closely together to get them executed.

Carlos González de Villaumbrosia | Product School 00:04:26 So engineering and product are peers, and now, what have I heard about design?

Carlos González de Villaumbrosia | Product School 00:04:31 Is design also a peer to the other two?

Saral Jain | Snapchat 00:04:33 Our design team reports to our head of product. So think of it as three in a pod. The design, product, and engineering teams work very closely together, but design and product report to the same leader.

Carlos González de Villaumbrosia | Product School 00:04:46 Got it. Well, one of the things I heard your CEO mention recently is that AI is writing over 50% of the code, and that designers are even shipping code. So, for you as an engineer, I'm curious, how do you feel about non-engineers shipping code, and how are you managing that?

Saral Jain | Snapchat 00:05:05 I have actually gotten very comfortable with that idea, because it is a reality in the industry right now that the boundaries between designers, product managers, and engineers are actually changing in front of our eyes right now.

Saral Jain | Snapchat 00:05:17 We have product-minded engineers working through the product process. We have designers, design engineers, and product managers shipping code in production, and that is the reality we have to live in, because the boundaries between these roles will continue to become less and less clear.

Saral Jain | Snapchat 00:05:36 And I don't think it matters what your title is. It really matters what you're spending your time doing, and that's really how we're approaching it. We definitely have some designers who are shipping production code. But we have over a decade of investment in our code base that makes it easy for people with non-engineering backgrounds to contribute code without increasing the blast radius of outages or performance regressions too much, because we've invested in a very robust platform under the covers.

Carlos González de Villaumbrosia | Product School 00:06:10 I'm always very curious about this topic. I'm seeing a lot of AI-native companies thinking about smaller pods. Anthropic is an example, I hosted their head of product recently, and in those pods, as you mentioned, regardless of title, everybody's shipping. So how does that work in practice for your org?

Startup Squads and Shipping Fast Without Breaking Things

Saral Jain | Snapchat 00:06:28 I think we have a similar concept. We call them startup squads, and this is something Evan published in his letter last year as well. He typically writes an annual letter to our community, to our employees, to our stakeholders. And in that letter, he mentioned this concept of startup squads.

Saral Jain | Snapchat 00:06:44 And these are typically very lean, cross-functional teams of engineers, designers, product managers, and data scientists, all working together on a problem, almost treating it as a large startup, a zero-to-one initiative, or a net-new revenue initiative that they're working very closely on. The role distinctions within those startup squads aren't as clear, because we're all working together as part of a startup to make something successful.

Saral Jain | Snapchat 00:07:10 And so, yes, there are designers who ship code. There are product-minded engineers who have a knack for asking not just the how but also the why, who are actually coming up with the product intuition as well. And the teams are working really well. I think we've invested in nine startup squads, and most of them have generated incredible results in a very short amount of time.

Carlos González de Villaumbrosia | Product School 00:07:33 And so from a quality assurance perspective, especially as more and more code is being generated by AI, and more and more people have access to shipping into production, how do you create the mechanisms to ensure that you maintain high quality as you increase speed?

Saral Jain | Snapchat 00:07:50 That is one of my top focuses right now, because, as you know, we ship to almost a billion people across the world, and quality is the utmost important thing, because we cannot afford for that to be degraded. The way we approach it is foundational platform investment. So we have a very, very good code review agent, for example, it's called CodePal.

Saral Jain | Snapchat 00:08:11 So no matter who ships the code, the first pass of code review is done by CodePal. It reviews more than ninety percent of the code within five minutes of the code being published, and it does a very, very good pass at identifying all the major issues that can happen. It understands not just the code that is being changed, but literally the tens of millions of lines of code across all of our repositories, all at once.

Saral Jain | Snapchat 00:08:34 So it can really identify patterns that humans might not be able to in many cases. And that's one good example of how we've invested in underlying platforms that can help us ship code faster across different job profiles without reducing quality. We have many such investments, like in our client platform, in our server architecture, where we are solving all of the hard problems once so that the feature engineering teams can focus on their feature logic.

Saral Jain | Snapchat 00:09:02 Does that make sense?

Carlos González de Villaumbrosia | Product School 00:09:04 So you're saying there's a coding agent that does the initial pass, and then you still have a human pass, where the engineers ultimately vouch for the code that goes live? Or is there anything else?

Saral Jain | Snapchat 00:09:17 I think that is where we are at today, which is, ultimately, every single piece of code needs to be owned by a human, because that ownership is extremely important for us.

Saral Jain | Snapchat 00:09:25 So the first pass is done by an AI agent, but then a human reviews the code manually. Although, in the future, we're getting increasingly more comfortable with, depending on the service or the kind of code, having the AI pass be the authoritative pass, although we're not completely there yet.

Saral Jain | Snapchat 00:09:42 And AI adoption will be uneven across a large company like ours anyway. But today, humans are absolutely responsible for the ultimate code that gets shipped.

Spinning Off Specs and Dotmo, and Outcompeting Giants 100X Your Size

Carlos González de Villaumbrosia | Product School 00:09:52 Got it. And so, on the org design front, I'm also curious about how you're now structuring teams across multiple companies, even, right?

Carlos González de Villaumbrosia | Product School 00:10:01 So you recently, well, earlier in 2026, spun off Specs, your AR division. And then, most recently, you also spun out another company, the gen AI video group, Dotmo. So we're curious to know, what's the rationale behind spinning off teams or companies instead of just keeping them as part of the same umbrella?

Saral Jain | Snapchat 00:10:25 That's a very good question. I think that's more business rationale than anything to do with product development or AI. But Specs is a great example. This is such an exciting moment for our company right now, we've invested in Specs for over a decade.

Saral Jain | Snapchat 00:10:41 We've invested in AR for over a decade, and things are converging, where AR and AI capabilities are strong enough, and the hardware is strong enough now, that we're close to our consumer launch. We announced Specs this year, and it's a different product. Snapchat is a mature platform used by a billion people.

Saral Jain | Snapchat 00:10:58 Specs is a net-new computing platform that's being done for the first time in the world. It's almost like a startup, right? And so those two require different kinds of thinking in terms of how to operate a business, different kinds of investments, and that's the rationale for spinning it off as a separate entity.

Saral Jain | Snapchat 00:11:16 Similarly, Dotmo, which is the new gen AI startup that we spun off, is a very creative idea, because we have almost a decade of investment in our own AR technologies, and we've been developing our own image and video generation models. We think we can continue to double down on those investments, which are very capital-intensive investments, as a separate entity versus part of the same umbrella.

Saral Jain | Snapchat 00:11:43 But they share a lot of technology together, right? As I mentioned, the foundations we've built over many years are shared across all of these. And so it's very exciting for me to see that we're able to support not just Snapchat, but multiple different products, using the same foundation.

Carlos González de Villaumbrosia | Product School 00:12:01 It's also very interesting to me, because outside of Meta or Google, Snap is kind of the independent social platform, right? The David fighting against the giants and somehow staying alive. So some of these moves you're making, like spinning off companies, I've seen them at much larger platforms, right?

Carlos González de Villaumbrosia | Product School 00:12:22 So I'm curious to know how you're able to really build your own infrastructure and manage your budget, assuming you have less budget than those giants, and still be able to stay relevant and competitive in such a market.

Saral Jain | Snapchat 00:12:36 I think about that so much. I believe Snapchat is one of the only independent platforms that has the kind of reach that we do, and that's been true for the entire history of Snapchat's existence.

Saral Jain | Snapchat 00:12:47 We've always competed against companies that have 100X more resources, 100X more budget, 100X more people to do it, and the only way we're able to do it is because it's such a creative company. We move so fast, and we understand our community so well. And those are part of the DNA of the company, and that has served us really well for the last decade, and I believe it'll continue serving us really well in the next decade.

Saral Jain | Snapchat 00:13:10 And especially with AI, I feel AI is the best thing that could have happened to a company like ours, which is just so creative, with so many ideas, and often the bottleneck is execution. But with AI, the execution bottleneck is going away, so it's a great time to be in our shoes right now.

Carlos González de Villaumbrosia | Product School 00:13:29 I want to double-click on that and bring the elephant into the room, right? Recently, Snapchat, or Snap, announced a layoff of around 1,000 people. And I put AI as one of the reasons, because you're finding more efficiency with it. So I'm really curious how you're able to still reduce your team, and yet maintain that crazy pace, and yet spin off companies.

Carlos González de Villaumbrosia | Product School 00:13:55 What's the secret sauce behind leveraging AI in ways that other companies aren't able to leverage?

Saral Jain | Snapchat 00:14:02 Yet? I think that restructuring was a reality of the business. And AI is one of the things that helps us feel confident that we're well positioned going forward.

Saral Jain | Snapchat 00:14:12 And we're using AI across many different aspects of the business. One is AI products for our community, for our Snapchatters who use Snapchat. Second is AI for our employees, how we're rethinking the way work is done within Snapchat and being more efficient using AI. And then the last thing is AI for our partners and advertisers, how we can rethink our business using AI.

Saral Jain | Snapchat 00:14:36 And we're making tremendous progress on all three of these verticals, and that gives us the confidence that AI can help us continue moving really fast. But I will say that, in general, I believe smaller teams are more flexible, more nimble, and able to move fast, and that's something that's proven really well for us over many years, and we'll continue doubling down on that.

AI Features Users Don't Even Realize Are AI

Carlos González de Villaumbrosia | Product School 00:14:58 So you mentioned there are three specific applications: one, external applications for the user in terms of AI; the internal application for your team; and then also for your ad partners. So I want to go a little deeper into each of those. Thinking about your external users, what are some of those AI features that maybe users don't even realize is AI, but is powering the experiences that people have on Snapchat and other companies?

Saral Jain | Snapchat 00:15:26 I think outside of ChatGPT and maybe NanoBanana, there haven't been a lot of breakout AI features that are massively adopted. And so, for us, we actually don't think of features as AI features or not-AI features. We think of features in terms of whether they're adding value to our community and our core product, right?

Saral Jain | Snapchat 00:15:46 And I think, in that, we've recently launched a couple of very interesting features that are gen AI features. One is My AI, which is our chatbot. Essentially, it's an AI assistant that people in our community use. But it's very unique, it has a personality, and very different kinds of use cases that people use My AI for.

Saral Jain | Snapchat 00:16:05 And similarly, we have gen AI lenses, generative AI lenses. AI lenses are a core part of Snapchat. Snapchat opens to the camera, so the first thing people play around with are those filters and lenses that are unique to Snapchat, and many of them are powered by generative AI right now.

Saral Jain | Snapchat 00:16:21 Last December, I believe, we launched a lens called Imagine Lens, which is an open-prompt lens. You can look at anything in the world, or your face on the camera, and give it an open text prompt saying, "Transform my world in this particular way," and it'll do it. And I think it's been such a massive hit with our community.

Saral Jain | Snapchat 00:16:40 I believe over 700 million people have interacted with our gen AI lenses since they launched, so, definitely, a big viral hit.

Carlos González de Villaumbrosia | Product School 00:16:48 You're right. A lot of these features, they don't need to be called AI features. At the end of the day, if it's a good experience for the user, it's a good experience.

Carlos González de Villaumbrosia | Product School 00:16:55 It doesn't really matter what's underneath it. And you're making me realize, I mean, Snapchat has been the pioneer for a lot of these new features, right? Since the days of Reels to now Lenses, to the location of your friends. A lot of the things that are now becoming commoditized, right?

Carlos González de Villaumbrosia | Product School 00:17:12 So where's the moat these days for those types of new experiences, especially as people are able to copy them faster?

Saral Jain | Snapchat 00:17:21 I think the moat is the platform and the network effect. We've always pioneered some of the most widely used features in the industry, whether it's ephemeral messaging, to Stories, to Maps, to gen AI lenses, to just filters.

Saral Jain | Snapchat 00:17:36 Snapchat's DNA is being creative, and I think our design team and product team are some of the best in the world. But we have a habit of doing it again and again. The most interesting thing is, the reason our community loves these features is that they're all interacting with each other.

Saral Jain | Snapchat 00:17:52 Ultimately, people come to Snapchat to talk to their close friends and family in a way that mimics the real world. It's fun, it's casual, it's not for likes. And I think that's why people love Snapchat, and we build features that all tie into this DNA of being able to talk to your close friends and family in a fun, unique way.

Saral Jain | Snapchat 00:18:14 And that's just how we think about product development, so that's our moat.

Carlos González de Villaumbrosia | Product School 00:18:18 That's true. The filters are another feature that you pioneered, and now it's assumed that everybody should have filters on camera. So let's talk about... you talk about close friends, you talk about family.

Segmenting a Billion Users, and AI for Advertiser ROI

Carlos González de Villaumbrosia | Product School 00:18:28 I want to talk about user segmentation, because, over the last 15 years, of course, there's been a lot of new users getting onto the platform. You mentioned you have over a billion monthly active users. So who is Snapchat for, and how are you making sure that you continue to be relevant to some of those new users, versus maybe other alternatives?

Saral Jain | Snapchat 00:18:49 Snapchat is for everybody. I think, in the past, there were conversations about whether a particular demographic was the right demographic for Snapchat. But when you're reaching a billion people, you can safely assume that everybody is your user, and that's true across geographical boundaries, across demographics, across age, gender, and everything.

Saral Jain | Snapchat 00:19:11 And our community is so engaged with our platform that it's one of the most often-opened apps in the world. So not only are people coming every day, they're opening it, on average, 30 times a day, right? And so it's a very repetitive behavior, and we love that about our platform.

Saral Jain | Snapchat 00:19:31 And so, no, it's for everybody. And we understand our user base really well, and our features resonate with a large user audience as well.

Carlos González de Villaumbrosia | Product School 00:19:39 I was having a similar conversation with the VP of product at Robinhood, right? They started as an easy trading app for first-time users getting into finance, and now, obviously, as they've evolved, they have sophisticated financial investors all the way down to Gen Z-ers who are also trying to get into the financial markets.

Carlos González de Villaumbrosia | Product School 00:19:57 Right, so how do you go about segmenting customers, or providing different experiences, to make sure you can really cover such a broad spectrum age-wise?

Saral Jain | Snapchat 00:20:07 I think it all starts with each feature: what is the value proposition? To give an example, we recently launched a feature called Topic Chats.

Saral Jain | Snapchat 00:20:17 Topic Chats is a place where our entire community can get together to talk about often-talked-about topics. And, most recently, we've realized that some of these, whether it's the NBA or the World Cup that's going on right now, have seen massive use of the Topic Chats functionality.

Saral Jain | Snapchat 00:20:35 And so, for us, it's about having a really good perspective on what we're launching and who we're launching it for, and then doubling down on building creative features to make it useful. But that segmentation, we try not to do ahead of time. We're thoughtful about which features will resonate with whom in the community, but sometimes our community surprises us, and so we then get creative and build features that can have the largest impact.

Carlos González de Villaumbrosia | Product School 00:21:00 So, shifting gears a little bit, I also want to talk about the AI use cases for your ad partners. I guess that's still the main part of your business model, right? Partners who want to advertise on Snapchat. So what are the specific AI-powered experiences for them?

Saral Jain | Snapchat 00:21:17 That is such an important thing in our business.

Saral Jain | Snapchat 00:21:21 Snapchat's ultimately an ads platform from a revenue perspective, although we do have a very thriving subscription product right now. But ads are the real way we make money, at least right now. Again, I think the way we approach AI for our community is the same way we apply AI for our advertisers.

Saral Jain | Snapchat 00:21:40 AI for AI's sake isn't something advertisers are interested in. They're interested in the performance of our ads. And so, ultimately, AI is only valuable for them when the ROI of every dollar they spend on Snapchat increases, and that's how we're using AI. So we've built solutions for smart targeting of customers, smart measurement of how ads are performing, smart creatives.

Saral Jain | Snapchat 00:22:03 So, how can advertisers create the creatives automatically, without too much investment on their side? So it's a pretty big investment, in all parts of the advertiser journey, using AI. But, ultimately, not for AI's sake, but just for improving the ROI for our advertisers.

Carlos González de Villaumbrosia | Product School 00:22:19 So, as I think about this, you're in the advertising business, you're competing for attention, ultimately. And what is the unique value prop, or how is AI helping you accelerate that value prop, so you can capture a bigger part of that wallet?

Saral Jain | Snapchat 00:22:39 I think we have to get advertisers to what the core moat of our product is, which is, for example, our messaging product, right?

Saral Jain | Snapchat 00:22:47 And so we've recently launched AI-sponsored snaps, which are sponsored snaps from our advertisers on the chat feed, but it's an AI-enabled product where users can interact with their favorite brands using AI conversations. And we recently piloted it with, I believe, Experian, for example.

Saral Jain | Snapchat 00:23:06 And we're seeing great lifts and great results, advertisers are seeing great results. And so, when we think about advertisers, we also have to tie it to our community and our usage of our app, and that AI is a value-add on top of it.

From Jobs to Be Done to Casper: How AI Turned Into Business Impact

Carlos González de Villaumbrosia | Product School 00:23:19 And then, finally, we already touched on internal use of AI, as you mentioned, smaller pods or squads, everybody shipping, creating the constraints for people to stay creative and use AI in ways that are helping you ship faster.

Carlos González de Villaumbrosia | Product School 00:23:35 But one of the challenges I'm seeing with organizations is that, as they build that system for everybody to be enabled to leverage AI, not everybody's on the same page. So you have this group of people who are super users, who are already creating their own context and really going for it.

Carlos González de Villaumbrosia | Product School 00:23:51 And then you have people who are curious, but they still need to be brought up to speed, right? So, as we get into a more multiplayer platform, with the idea that everybody should be able to leverage AI without having to set up the entire system or context themselves, how are you building that internal infrastructure?

Saral Jain | Snapchat 00:24:10 I'll talk about engineering, and I'll also talk about the larger company, because we're seeing different things in different use cases. First, talking about the company: when we first started rolling out some of these AI solutions, like Gemini and ChatGPT Enterprise, we quickly realized that almost everybody was using it on a daily basis, so adoption wasn't the problem.

Saral Jain | Snapchat 00:24:30 Translating that adoption to business impact was the problem. A lot of prototypes were being created, and many of them were being thrown away, and that's not necessarily useful for the business. And so we quickly pivoted to what we call the jobs-to-be-done framework. We actually sat down as a leadership group and talked about every function, whether it's sales, engineering, marketing, legal, or any other function in the company.

Saral Jain | Snapchat 00:24:55 What are the actual jobs they do for their stakeholders or users, right? Whether it's internal employees, external Snapchatters, or anybody else. And after we identified the jobs, we then started talking about how AI can help us rethink the job, accelerate the job, or maybe not touch that particular job.

Saral Jain | Snapchat 00:25:15 I think once we created an operating system that's oriented toward the jobs we need to do for our community, then AI became a massive accelerant. And, of course, even there, we had to go through a change management process of training and enablement, and making sure we were building the agents that could help us accelerate the jobs.

Saral Jain | Snapchat 00:25:33 But crafting and grounding things into the jobs-to-be-done framework was extremely helpful for us.

Carlos González de Villaumbrosia | Product School 00:25:39 How do people use that framework in practice?

Saral Jain | Snapchat 00:25:42 In practice, we actually created a list of our jobs to be done per functional area, and then, when we create our roadmaps for the startup squads, or generally for the entire team, we align the roadmaps to the jobs to be done.

Saral Jain | Snapchat 00:25:54 So, at any given point in time, you can actually see what the jobs to be done across the business are, and what roadmap items any individual is working on, and which jobs to be done they're tied to. And so, when they actually start using AI, instead of working on some throwaway prototype, they use AI for one of the actual roadmap items that's tied to the jobs to be done, and that's how business impact happens.

Carlos González de Villaumbrosia | Product School 00:26:17 Got it. So it helps you put the guardrails around what the company's prioritizing, and I guess you also have the infrastructure in a way that allows people to start building with AI without having to do all the back-end work. I was so curious, you mentioned at the beginning that you have a coding agent that helps do the initial pass for code, and...

Carlos González de Villaumbrosia | Product School 00:26:38 But you also mentioned there are other, non-coding teams outside engineering that are also leveraging AI. So how are you thinking about those agents? Is it that you have a generic agent that's overseeing everything, or is it more specific agents, divided by function?

Saral Jain | Snapchat 00:26:53 So we have two kinds of agents. Some are foundational agents, so a code review agent or a coding agent are foundational agents that anybody in the company can use.

Saral Jain | Snapchat 00:27:17 But then some are vertical agents as well. We have an agent that does a great job at A/B tests, it looks at all of our A/B test data, experimentation data, and is able to understand what's going on.

Saral Jain | Snapchat 00:27:37 We have a data science agent. We have a cost agent, which finds cost inefficiencies in our infrastructure. So we have vertical agents and horizontal agents. One of the best examples of a horizontal agent is called Casper. Casper is our remote coding agent. So what it does is, we almost call it an AI teammate.

Saral Jain | Snapchat 00:28:10 It's not an engineering agent, any team in the company can use Casper. And Casper is available where people already are. So it's in Slack, it's in Jira, it's in Coda, other tools that we use. And at any point, you're having a conversation with your teammate, you're just thinking about a product idea or some iteration, Casper is listening in on that conversation, and at one point, you can just invoke Casper, saying, "Okay, you have all the context, you know everything about the history of the company, and all of our knowledge base, and literally all of our tens of millions of lines of code across all repos. Why don't you go ahead and build it?" And then Casper goes ahead and builds it. And then the code review agent, which is CodePal, reviews the code, and they have a back-and-forth, and that's how engineering is working right now. It's a super wild time to be an engineer.

Carlos González de Villaumbrosia | Product School 00:28:23 That's so cool.

Carlos González de Villaumbrosia | Product School 00:28:24 The same way we talked about designers now shipping code, I also want to ask you about engineers prototyping and getting into the designer territory. How do you go about it now?

Saral Jain | Snapchat 00:28:35 That's actually such a refreshing change from my perspective as well. Not that we didn't have that in the past, but we always look for engineers that are product-minded, because, ultimately, it's a product we're shipping to our consumers, right? And so we always look for traits in an engineer where they're not just asking how to build something, but why something should be built, and that trait goes a long way in building prototypes, as well as building actual products for our community.

Saral Jain | Snapchat 00:29:03 But we're seeing so many examples of engineers coming up with insane ideas, and now the bottleneck isn't Figma mockups, the bottleneck is we just build the prototype, show the prototype, and the best idea wins.

Carlos González de Villaumbrosia | Product School 00:29:14 So, as you build those prototypes, are you still relying on third-party tools like Figma, or do you go straight into Claude Code and prototype?

Saral Jain | Snapchat 00:29:24 Most of the prototyping is being done in Claude Code right now, but we do have our own design system that Claude Code understands, and so it still looks and feels like Snapchat. And, of course, Figma does play a big role in terms of ultimately shipping the features that we ship, but prototyping is being done, feature by feature, by tools like Claude Code, Codex.

Saral Jain | Snapchat 00:29:42 We've rolled out every AI tool you can possibly imagine. And so I wouldn't necessarily single out Claude Code, but certainly Claude Code is a big part of it.

Why the Market Lags Execution

Carlos González de Villaumbrosia | Product School 00:29:50 All right. No, and I find that really cool, because you're literally setting up the conditions for, in this case, an engineer to be able to prototype, and the design system is there, the tooling is there.

Carlos González de Villaumbrosia | Product School 00:30:01 You can go straight to the actual experience you want to create, and vice versa, right? When a non-engineer is trying to ship code, if the conditions are there, they can go straight and request and try to do what they want, without having to worry too much about all these other things that might be more technical.

Saral Jain | Snapchat 00:30:17 I think we want to get to a world, and we are in this world, where, in a scenario when building is so cheap, what to build becomes the most interesting thing, right? And that's where the best ideas win, and the best way to find the best ideas is to have lots of them. I think Evan has talked about this quite a lot as well, and that's where prototypes come in pretty handy.

Carlos González de Villaumbrosia | Product School 00:30:38 And so, as you continue to make these big bets, spinning off companies, leveraging AI in a crazy way to build faster and better, then there's also the reality of the market, right? Kind of like the weather, you check the ticker and it's like, oh my God. Why isn't the market, in your opinion, rewarding all of those investments and improvements you're making, even though you're objectively beating expectations across the board?

Saral Jain | Snapchat 00:31:05 The market is sometimes a lagging indicator of execution. And so we're heads-down, focused on execution, executing our strategy. That's something Snapchat has done really well for over 15 years now, and we'll continue doing that. We know, for a fact, that our community loves us, that's why they use...

Saral Jain | Snapchat 00:31:21 Almost a billion people use us every day. We know our advertising performance keeps improving, and we know that we're focused on being disciplined about our gross margins. And so we know the investors will eventually catch on to that, but that's not something you, as an engineering leader, want to distract your team with on a day-to-day basis.

Carlos González de Villaumbrosia | Product School 00:31:39 For you, from a planning perspective, how far do you try to plan, knowing that there are so many changes?

Saral Jain | Snapchat 00:31:46 In the past, we used to do these year-long plans, and we still have a sense of what our strategy is for the next year. But we try to break down planning on a quarterly level right now, so that we have clarity on what we want to achieve every quarter, tied to our yearly plan.

Saral Jain | Snapchat 00:32:04 And then, finally, we're extremely flexible, and things change on a daily basis. We sometimes actually do daily check-ins on key initiatives. And, as things change, we're adaptable, because that's in our DNA, to be able to move fast and be adaptable and flexible on things. So I wouldn't recommend, especially with AI, doing these long-horizon plans and sticking to them, because that's just not the reality of the world at this point.

Carlos González de Villaumbrosia | Product School 00:32:24 Right, and it's so part of your DNA. You're the ultimate survivor, literally 15 years fighting against giants, and innovating, and finding creative ways to continue the good fight. Saral, it's been a pleasure to learn from the inside how you guys are building and thinking about AI.

Carlos González de Villaumbrosia | Product School 00:32:44 Thank you so much for going deep into the details with me.

Saral Jain | Snapchat 00:32:47 Thank you. That was such a fun conversation.