What the AI Boom Means for Databases and Enterprise Software

0:00I'm fresh off the mongodb.local event. They had a ton of developers, partners, startup founders hanging around and showing how Mongo is trying to respond to everything that's happening in artificial intelligence. They are the sponsor of this podcast, and they had me on site at mongodb.local to do a bunch of these interviews, which I really enjoyed. So we decided to put them in our feed. I had a great conversation with CJ Desai, the CEO of Mongo, who dropped by my little studio at the conference. We have another interview with the head of AI from Rippling coming after that. This is the Newcomer Podcast.

0:43Hi, I'm Eric Newcomer, author of Newcomer. We're here at MongoDB's local event. The amazing drop-in when you get the CEO, what, 65 days into the job to show up on stage. What, I mean, this is your big event. And what's the message that you've really wanted to carry to the attendees here? You know, I would say, Eric, first, thank you for having me. And thank you to you for letting me crash the party. Of course, I love it. Literally, we just talked to a three-person company. So now we're going much larger. Yes, sounds good, Eric. So 65 days in, and one of the things that our previous CEO, Dev, and the entire team, we realized, so first of all, from just being in Silicon Valley for a long time, San Francisco feels like it's back.

1:41San Francisco is back. During the pandemic, people went a little dark on San Francisco, but with this AI platform shift, San Francisco is back. and as you know, we are a New York headquartered, New York founded company. Listen, I'm a New York based person who also believes in San Francisco, so we have a shared spirit on that, realizing that AI boom is here, we're also in New York, so yeah, I feel it. So, San Francisco is back. MongoDB, about 10 years ago, did a really nice job with San Francisco. They were in front of software developers, builders saying build on MongoDB, here is why, speed, agility, scale out, many, many advantages.

2:26And then the team realized, Dave and the team, that we kind of, as the company succeeded, we took our eye off the ball, for lack of a better term. And so after four years, we decided to reintroduce MongoDB today in San Francisco. So today is an auspicious day. Thank you for coming. and we wanted to launch and tell people that we are here. We are MongoDB. It's a great data platform you can build on no matter what kind of applications you are building on, digital, native, AI native, or you have AI++, whatever the case might be, it is a great database. And one of my biggest profound realization was that when I was doing my own diligence to join MongoDB, even though the founders didn't create with AI in mind, unstructured data, flexible schema, semantic retrieval and all these things that are part of now the platform, it is like the platform for AI applications.

3:30And so we wanted to tell that to everyone and we got a great customer today to validate a great founder who said he has built three companies on MongoDB. You're talking about Mike Krieger. Yeah, you had the nice self video Instagram founder. I created Artifact, which you talked about briefly, which is an interesting for me in the news world. And then obviously he's at Anthropic. Yeah, that was an exciting endorsement. What is your read on, you know, I don't know, AI hype in 2026? On some level, you must think this is obviously a phenomenon to say you're doing this big San Francisco push. Dave, you know, I talked to him last year.

4:10You know, he has a dose of realism to it all. What's your personal perspective on where we are in this hype cycle? He's smart and realist, and I'm an optimist. That's how I would... You brought in the guy like, oh, no, this thing is going strong, and we're going to get in front of it. Yeah, I mean, I would say since 2022, Christmas-ish or October-ish, when you look at the evolution of AI, and you look at some of the killer AI companies where they have scaled business in a significant way, has been, of course, OpenAI. Everybody understands that that is a killer app from my perspective. But you also look at Grok and how fast they have grown XAI.

4:52Right. And then Anthropic, the entire team at Anthropic Lab, that is a truly killer AI app for coders. I mean, people love that platform and they sometimes use it with cursor, this, that. So you are seeing that these companies, if you look at like internet age in 90s, you look at mobile age, they did not scale this fast. How fast these companies have scaled, that's like pretty amazing from my standpoint. And in 2025, I think the killer apps were the coding tools. They were the killer app, besides ChatGP, of course. And now as we go into 2026, which ones are going to take off? I'm optimistic that some very specialized vertical apps related to maybe a healthcare or insurance assist or something may take off.

5:45But you will see some take off. We're close to a bridge and open evidence. Those are super interesting applications in health. You interviewed Constantine and Sequoia. Obviously, they're bullish on Harvey. So a lot of these vertical applications are exciting. And Harvey, I met them a couple of years ago. and they were still, you know, truly with the legal firms, they were using them. And then now you have the in-house counsels use them as well. And so they have gotten a perfect product market fit and they're expanding now their use cases. So that's just it. But that's very specialized, right? It's not just generic.

6:25I come in and help me review this contract from a legality perspective and so on. This is a drop by, so I don't want to take too much of your time. This will be the last question. How do you think about, you know, there's so many models available. How do you think about like, oh, where to provide your own versus to just sort of say, you're obviously going to bring models from other places. We provide sort of the database layer. For databases, we want to be model agnostic. Yeah. We want to provide best embeddings. So your retrieval quality is high, accuracy is high. But we want to be model agnostic.

7:01and when I speak to customers, including AI companies, like everybody originally was telling me open AI, then they shifted to cloud. I think the innovation cycles are very fast and then you look at some of the firms in China and you look at DeepSeek and others, that innovation cycle is all... Right, it can come from anywhere. Yeah, that could come from... And the product cycles are shorter, right? Product cycles are shorter. So multi-model, not multi-model, but multi-model will be a way to go. And it gives freedom to people to use whichever model is best. Gemini maybe today better. Tomorrow maybe on topic again.

7:40And I think that's how it's going to work. So we will be agnostic regardless of who you use. Our goal would be always to give best semantic retrieval capabilities and completely provide a scalable data plan. Well, I love, you know, Newcomer, we use green. So it was easy to share the stage, but honored that we get to share branding with Mongo. Thanks so much for having me here. And thanks for dropping by our stage. We appreciate it. And we'll see you in Brooklyn soon. Sounds good. All right. Thank you. Thanks again to MongoDB for sponsoring this episode. I feel like I've talked to both CEOs, the old and the new, in the last six months.

8:17So that's been a lot of fun. And now excited to get into it with Anchor Bot, the head of AI at Rippling. I know his big boss, Parker Conrad, pretty well. He's been on this podcast before. Parker has been somewhat slow to embrace AI. So it was fun to talk to his head of AI about, you know, the company finding AI Jesus, believing in AI a little bit and how they're making that cultural and technological transformation. Give it a listen. Thrilled to have Anchor bought the head of AI at Rippling here, I was saying before we got on. I go way back with Parker Conrad, your CEO. I wrote about Zenefits back in the day.

8:52I feel like I was, you know, as far as reporters can be bullish, I was pretty bullish on his comeback in Rippling and have been following your work. What's it like to be the head of AI for, I feel like for a technology, Parker is a technologist, forward thinking guy, but he has not been wrapping himself in AI. Like, how is that? How is it to be the head of AI and CEO sort of coming to the AI hype? I think it's been exciting, specifically from the lens of there is so much happening in AI every day. So automatically, everybody's curiosity on what's the new model release, what's the new capability release.

9:31Like these days, everybody is hyped up on cloud code, as an example, what Enthropic is doing. So that automatically starts to create this pull internally around what does this mean for rippling from a product point of view? what does it mean for rippling in terms of day-to-day business perspective and what does it mean in terms of customers expectation on rippling right you get you get to be the guy who knows what's happening to ai it's like oh how do we respond to this this that and the other we're obviously you know at mongodb.local uh you know obviously sprawling ambition with rippling it's like oh we want to stack startup on startup so i can imagine uh strong data organization capability is a part of the business success but explain rippling's relationship with mongo yeah so we have been a long time customer of mongo uh from the beginning i think as you rightly pointed out parker's thesis was a compound startup like product over product and the core heart of it is the employee graph and being even though yes people know us that we can run payroll for you core of it is the employee graph which captures not just your pay related information or benefits related information we also capture your it and identity related information your device related information because at the end we also have a product portfolio of it similarly we also have a product portfolio around finance your corporate card your spend your travel expense we just launched a travel expense product.

11:08So that suddenly, the amount of data you have about employees gets connected in so many different ways. And having a partnership with Mongo allows us to continue keeping that graph sanity intact in terms of the relationships you have in this business data. How rigorous are you guys about making sure that everything built at Rippling uses the same tech stack? I mean, if you're acquiring startups and you're letting people sort of do their thing, how do you balance the trade-off of sort of a consistent stack versus letting different people build what they want? Oh, that's a great question. And speed versus consistency is a constant dialogue we have internally.

11:50I think there is a value in terms of letting people innovate on the side, but it's a little bit of a trade-off because as soon as they are reusing the common capabilities we already have in our platform, there is a dimension of speed which comes with it. And suddenly, for example, the employee graph comes with the workflow. It comes with the notification. It comes with the inability to track the changes that's happening. Now, if I am building now a travel expense product, there are workflows for me. Of course, there is a dimension where I need to go into integrating into flight booking, hotel booking, which is very unique to my products, you know, niche.

12:36And there I'm free to choose what I choose to do that and run faster. But I think we've been able to guide the different product teams, even though running independently on the core value of building on that common employee graph and common capabilities. What has been the value of Mongo? Would Rippling be a different company if it was built on Postgres? I think it would, no, Rippling would not be a different company because of Parker's overall hypothesis of a compound startup and really building it as a connected ecosystem of products running on a common foundation. I think it would have been just a little bit harder.

13:18And having that partnership with Mongo from day one allowed us to build this employee graph, build a capability set so that the next product or the next product we were creating was a lot more easier. And this is not just about transactional data which we are maintaining and managing. It's also about analytics and reporting, being able to take that information and have ability to report across a cross section of your product portfolio or your business data, not just from HR or IT or finance, allowed, has opened up those possibilities, which having a common data store, having this Mongo capabilities, opened up those possibilities for us, which we would have solved eventually.

14:06It would just have taken a lot more effort. So you're the head of AI. We touched on some of that role means translating what's happening in the broader San Francisco ecosystem to Rippling. But in terms of products within Rippling, what are the AI products? Do you have an agent yet? What's that look like inside of Rippling? Great question. I think originally Rippling also started, we have almost 70 plus products in the portfolio. and originally each product started innovating around how they can embed AI as they're building that product, right? It could be within our recruiting product portfolio.

14:43When you're doing interviews, you are meeting candidates, summarizing those information. It could be within our IT product portfolio where you are issuing devices, you are tracking compliance and security around those things and having a summary of your policies. So people started embedding AI as product capabilities very early in ad-ripling. I think the change happened at the point in time we actively started investing on for building AI products, you have to use AI every day. Without that, I genuinely believe it's very hard for people to reimagine what it means for their product. So 10 months ago, Albert, our CTO, really put a charter out and that's where I came in, is driving the AI transformation of Rippling internally.

15:32That we use AI every day. We embrace latest and greatest of AI tools every day in our workflow, whether it is product, design, engineering, legal, finance, sales, marketing. And that... Is that Cursor, Harvey, Sierra? What do you use? Oh, the tech stack is full of all the tools you can think of. So, of course, Gemini, ChatGBT, Cursor, cloud code codex so access let engineers sort of pick their preferred i i think so because rippling deals with a lot of enterprise customer data we do have a ai pilot process and i have built up a checklist with our security and legal so that we can quickly assess a new ai solution somebody will request so i get request every week okay i want to try granola i want to try whisper flow typically what we do is we run through it through an ai pilot process you know do you use factory i'm going to be talking to the factory see you know it's on my list of from somebody maybe it'll get accelerated yes that could be i'm very keen to learn more which keeps my job a little right curious because a lot of times i can't stay on top of things so then the rest of the company is constantly keeping top of things and asking me can we try this or can we try that.

16:50And do you have an overall agent at Rippling? That's one of the new investments we are making. I think Mongo's partnership has helped us because if you think about Rippling, we are not just a payroll system. We are not just a finance or a travel expense system. We are a full suite of products. So for us, we can build an agent which is not a payroll agent or a finance agent or an IT agent. We can actually build a Rippling intelligence agent which can actually answer questions across your entire day-to-day business. And that's essentially what we have embarked upon. Parker is extremely excited about.

17:29Or business intelligence, you think, is a key output? I think it is really productivity. So if you think about AI, productivity is the impact which we see in engineering or product and design or marketing. The same productivity impact is what we are hoping to provide to our customers. because as our footprint has increased, our customers' growth has also happened. So they are growing from, like look at Android, one of our flagship customers. As Android is growing, they are using more products when we're playing. They have larger set of toy base. So essentially, their day-to-day operations has gotten complex.

18:10So now, if we are able to offer them an agent to operate Android payroll to their IT, to their travel and the agent takes care of, you know, proactively working with their administrators and making sure Android stays on top of things which they need to stay on is the value we are aiming for. You know, I hate to bring up a competitor, but, you know, Ramp and Rippling start in different places. They overlap on some, you know, you now have, I think, credit cards. Yes. And you both want to be sort of the, what, what's the term of art? like the record of all things business for some reason. They've clearly leaned into the AI brand more than you have.

18:53I don't know. What do you, what do you have? You're the AI guy. Like, do you want, do you want that AI brand or it's like we have different sensibilities or how do you think about it? I think there is a, there is obviously an equity in terms of having a clear AI story around your company and your product because customers are keen to understand that specifically because if I'm a Rippling customer today or I'm a prospect evaluating Rippling today, I come with that expectation that there is a certain amount of AI automation, productivity I will get. So I commend Ramp for what they have been able to embark on, not just on the product side, but as well as in terms of positioning themselves as an AI forward company.

19:39I think Rippling is not very far in terms of our AI transformation internally, we are very further along than what people may be aware of. In terms of our design partnerships, we have design partnerships with Cursor, with OpenAI, with Anthropic, with AWS, where we get early access to the latest and greatest Langchain, Databricks, and even Mongo. So early access to their product capabilities. We are piloting that at Ripling, using that to create productivity and give feedback back. And when I see those interactions and when I see our ability to influence product roadmap of so many ecosystem of AI companies, it reflects back on Rippling's AI impact is much larger than what people may be aware of.

20:24I should know, is Enterprise Search part of Rippling's vision? Because that's been an area, you know, obviously Glean is a very promising AI company. I would say part of Rippling's AI assistant will take care of answering questions you'll need answers for across your enterprise, being able to obviously, we being the system of record of a lot of information, you don't need glean in those cases because we can just answer those questions for you directly. Right. What do you think, you know, there's been a lot of like, AI is going to replace software companies. I mean, so you can get it from both ends, embrace AI, but also like, oh, you know, individual companies are just going to spin up their own payroll because, you know, They can build it with cursor on their own.

21:11Like, I don't know. Payroll, obviously, to me, seems sort of absurd because there's all the regulatory and it's just like sensitive. But clearly there must be pieces of your business where people are sort of hacking together their own apps. Like, what do you make of this narrative that the rise of artificial intelligence is going to have all these sort of homespun software applications? We actually embrace that wholeheartedly. because what we find at Rippling is, what we find at Rippling is as customers are using our platform, there are always niche, unique use cases for which they like parts of what we offer to them, but they want to extend, enhance, and add new capabilities on top of Rippling, which is why beginning of last year, mid of last year, we launched our custom app capabilities where you can essentially wipe code an extension app to Rippling at Rippling's platform itself.

22:08And again, going back to since we are at Mongo's conference, being on Mongo in terms of ability to store generic artifacts and documents actually opened up that possibility for us essentially then to create custom objects which then can be used as a container for any customer to bring in whatever data they want to bring in. Any interesting examples? Oh, tons of very unique, very interesting examples in companies because we deal with a lot of workers who are dealing with shifts, for example. And compliance comes up quite often in terms of having them, having gone through a certain amount of rigors in certain industries.

22:51Very unique cases in terms of tracking, you know, ticketing, tracking, signing sign outs, tracking. so there are like niche sort of use cases people are building which has actually motivated rippling to even launch our fd team which is actually now going in to our customers and building these niche custom apps for them and then seeing how we bring that into our platform as capabilities in terms of enhancing our platform to make that easier going forward do you think you're not your code base gets a little worse the more people use ai tools or what what's the risk in terms of over reliance on cursor that's a great that's a great question i think i think we've started to observe that yes there is a there is a dimension of an ai slop seeping in to the engineering discipline of coding every day right but i think we've been very strict about our ai stance on couple of things.

23:57First, testing is non-negotiable. So any code which is being checked into the main line and pushed to production has to be fully tested and vetted. Second, accountability. Just the fact that I used Cursor in AI doesn't take away my accountability of what that code does. And I think we've been very clear to engineers from day one that accountability is still theirs. So that essentially creates sort of a responsibility, accountability in engineers to be cautious of it. But that's just on the human side. From a system side, we're also bringing in additional AI tools. So AI coding, code review tools, AI tools around looking through in terms of production outages and bringing that information back to engineers to be able to troubleshoot and then improve.

24:46Because at the end of the day, the AI enablement or AI use is not just about development. AI can also be used in analysis of what's happening in production. AI can also be used in reviewing and verifying. And essentially, by bringing those systems in those places, overall keeps the health of the code base at a peak and avoids the slops sort of just bleeding into production. In terms of the rippling product, have you started to build it at all with the idea that it's like someday soon people are going to have AI agents that they treat as a worker. It's like, oh yeah, I want to in some ways compare apples to apple, this human and this AI agent, and I'm going to like think about it in terms of identity in the same way I would use Rippling.

25:39Has that started to creep into your... Great. I don't know. It's like you're just sort of picking up things we are discussing internally. Because one of the things we've been thinking about is agent identity, right? Because now if we think about agents we are building for our customers, at the end of the day, they'll start accessing payroll information. They'll start accessing information which has its own access management of who can see what and how and what type of actions they can take. So it does need three things. One, obviously an identity of an agent, which we know who's doing what. But then second, it should still inherit from a human.

26:21because in some ways that's one of the conclusions at least we landed on that at the end of the day who's culpable for this agent? Exactly, exactly, exactly. Who is on the hook for the agent at the end of the day and whose team this essentially which translates into whose permissions, whose access management this agent has to have and we found that it's a lot more easier to say that okay there is always a human there is always a human in the system of course what we are adding is a productivity element so instead of in future thinking of adding more humans you are essentially adding more agents but they still function as part of a payroll team and then they get access according to being a payroll admin they are part of an IT team and they get access as part of an IT team I feel like a lot of people are excited about what agents can do but not excited about being the one responsible for stopping the agent from doing what it's not supposed to do Is Rippling willing to take that on at all?

27:20Or do you see a world where you say, yes, we're the barrier for your agent not getting access to some system it's not supposed to have? That already is something part of Rippling's ethos. So because we are not just a payroll company, we also have a full-blown identity system. We have a full-blown access management system. We already control what data who has access to. We have a very well-defined access management platform, essentially. And if you, which is what is allowing us to now bring in agents and say, oh, this agent inherits from Ankur. And whatever Ankur can do in Rippling, this agent can do.

27:58Right. And that philosophy essentially allows us to guardrail, not just when we are building agents ourselves. Right. But customers deploying agents in their landscape, being able to then start performing tasks and actions and leading to unintended consequences. Right. Because we will still keep that principle even for them accessing rippling data. We guardrail that heavily the API access is only enabled through certain controls and access management principles. So randomly customers can't deploy an agent and start accessing rippling information. Well, you have what every AI company wants, which is great data.

28:37And now it sounds like you've got the AI conviction. So exciting times ahead at Rippling. And thanks for joining us. No, thanks for the great conversation and really enjoyed talking to you. And yeah, let's continue this dialogue. Sounds good. Thanks so much. Thank you. Thank you for tuning in to this week's episode of the podcast. If you're new here, please like and subscribe. It really helps out the channel. Listen in for new episodes every week, wherever you get your podcasts.