MongoDB CEO CJ Desai on 60,000 customers, AI-native growth reacceleration, and M&A strategy

waiting any longer. Thank you so much for taking the time.

What a set.

Jump on the show. How are you doing?

I am doing fantastic, man. It's like there are

700 people here. San Francisco is back. I feel like San Francisco is backing and we are re reintroducing MongoDB in San Francisco today. So, Phils, the energy is just amazing.

Fantastic. Uh, congratulations. Take us through some of the announcements. What does it mean to reintroduce MongoDB? Uh what are the misconceptions where people might think of it as one thing and now you need to reintroduce it?

I would say you know MongoDB first of all database market

has been around for a long time 50 plus years maybe 60 years.

Oracle is going to celebrate its 50th anniversary

year and a half. That was my first job out of college is I joined Oracle. Yeah.

And MongoDB

was a truly disruptive database that got created

in 2007 and in New York City. I mean this is a one of the few tech companies that got created in New York City

and put it on the map.

I know. I know. And then the leaf you know people ask me what is the leaf of MongoDB all about? M

and the founder said, "Hey, this is a natural way a software engineer or a builder should interact with a database versus a rigid way of interacting with database." So, it's pretty cool. I just became CEO couple of months ago. Uh came from Cloudflare and I'm having great time meeting innovators, customers around the world. on the misconception here's what I would say

that even though the database was created in 2007 the whole idea behind it was speed agility and dealing with a lot of unstructured data like emails [clears throat] PDFs images log files this and that so when the founder created this database it was not with AI in mind and today when you look at any AI application your favorite uh consumer application you may be using on a daily basis. You see there is a images in there. You may have create a video in there. Lots of interaction. This database was created

for the AI wave and we want builders building on MongoDB as they create agentic AI applications or any applications for that matter that has structured data so that they can move fast and hopefully create the big iconic businesses legendary businesses in Silicon Valley on top of MongoDB. Where are you seeing the most uh excitement and growth in your business? Uh I can imagine that there's a ton of green field projects that are happening in large clients, large enterprises where they want to move even faster. So they're grabbing MongoDB in addition to other modern tools on a new green field project. And then there's also a new startup boom of AI enabled software companies. Uh I'm sure that's growing as well. Uh what have you been seeing? Where's uptake the fastest in the last few years?

I would say John, you know what I learned and I can feel it right here on the floor is that you had pretty much customers. We have 60,000 customers. Okay. 62,500.

And our

That's fantastic.

And our customer growth Yeah. has accelerated in last 12 months which is pretty cool because lot of developers are now building on MongoDB.

Yeah.

Nice. I like it.

Every time you say something great I'm going to I'm going to hit that. So fair warning.

All right. I love it. So one is our customer base reacelerated as developers are building on MongoDB. But it's a tale of two cities right? One is large enterprises John as you called out they are growing still with MongoDB. They have some mission critical across every industry. financials, healthcare, government, I can tell you many across the globe just building on MongoDB. And you have AI native companies or digital native companies whether they are in San Francisco, Seattle, Tel Aviv or even New York.

They are building on MongoDB. So the customer growth is coming from them. And customer growth accelerated in last 12 months for us which is really cool. many many companies, you know, when they start deselling, they do not reacelerate and we reacelerate it. So that gets me like really fired up and really positive and the innovation that we announced today that if you want to truly build an agentic system at scale that doesn't hallucinate you can build it on your laptop. You always require a great database. You need an awesome agentic rag or you know semantic search that just works that is accurate. And this is truly a mission critical app that an insurance company may float it to their customers or a government may float it to their citizens or whatever the case might be. We want to make sure that this is a great data platform. So we announced voice for embeddings.

Uh we created a small embedding model called nano which is 1/8 the size of large. M it shares the embedding space and works with pretty much on cloud onrem multicloud and you know Jensen talked about at CES the whole thing around multicloud multimodal and multimodels and we do the same so very cool excited we gave that message this morning uh to the builders here we have a full show all day today which was also live and people are fired up

how are you thinking about uh M&A this year. Obviously, the company's doing really well. You're public company. You have a public currency for acquisition. Uh there's a lot of different pieces in the landscape that could be synergistic, but uh what's your thought process for uh partnering with uh other startups either uh in a full acquisition or another partnership? How are you thinking about it this year?

So, John, John, I'm a big believer in organic growth. Mhm.

I like we have a large market that we participate in.

We have only penetrated 2 and a half% of that market. 100 billion we are at 2 and a half billion.

So we have a lot of room with our core data platform. Wow.

Now we are always looking

for specialized skills say maybe in the AI area. So in 2025 we bought Voyage which was created out of Stanford. Uh amazing uh people you know the founder teaches a legendary course on machine learning. So we bought that company and he got amazing amazing engineers and researchers and that professor is now chief AI scientist at MongoDB. So we will do those type of acquisitions where they are small uh either for a technology or people or sometimes both

so that we can truly leverage and accelerate the pace of innovation.

Uh has anything in the uh the various compute bottlenecks been a headache or kept you up at night? I mean, we're reading about Amazon Web Services is paying for copper refining because they need to go so deep in the supply chain that they need to go to the mining level. Uh memory prices are through the roof. Uh there's delays on all sorts of custom silicon chips. Uh I know uh you might not be affected, but how are you thinking about the uh the the the different bottlenecks in the cloud computing uh supply chain? I would say the bottlenecks if you remember a couple of years ago everybody was worried about will they get enough GPUs from Jensen. Everybody's like, "Oh my god, bro. Are you going to get enough GPUs from Jensen?" Yeah. And uh that was like a big worry. Yeah.

And now when we speak to customers,

that's why they are like okay will GCP or Amazon like you said AWS and others will they have enough capacity? right now it is not something that people are panicking and Amazon or Google I would say they're trying to stay ahead on that curve so that it doesn't become that hey CJ we cannot run MongoDB for your favorite customer X because we are out of capacity man we just can't do that so we are trying to definitely uh work with them understand how they are dealing with both compute as in CPUs as well as GPUs and those innovation cycles are also very asked, but right now we are relying on them to solve, but we are not worried about it because they haven't given us the signal that we won't be able to do this.

Well, congratulations on the event. Jord, do you have anything else?

And and and the new and the new role. I I know I know we're out of time, but uh I'm I'm sure

I'd love to dig more into your background and and tell the fuller story. Uh but we know you have a busy day today. Congratulations and we will talk to you soon.

Give our best to everybody at the event.

Thank you, Jordy and John. Appreciate it.

Have a great rest of your day. Goodbye. Cheers.

Um, Lambda, Lambda is the super intelligence cloud building AI supercomputers for training and inference that scale from one GPU to hundreds of thousands. Uh, we have Matthew Prince from Cloudflare joining in just a few minutes. Uh, in the meantime, we will go back to the timeline. There is some massive news.

I did on on merge. I wanted to say a couple things. So uh male who's one of the co-founders emerged says TLDDR we are developing a new paradigm for BCI using molecules instead of electrodes.

Oh interesting.

Uh if you're excited about this and want to contribute in protein engineering synthetic bio delivery immunology ultrasound devices neuroscience or data MLAI we'd love to hear from you. So, uh,

that sounds extremely sucky.

Merge is, uh, pursuing ways to make BCI much higher bandwidth, millions of neurons, and less invasive, no device implants.

Instead of electrodes, we're focusing on modalities like ultrasound that can cover large areas of the brain

and on proteins that can connect these modalities with neural activities.

Uh, and so again, uh, different, uh, quite a different approach than Neuralink. Well, let me tell you about Turbo Puffer. Serverless vector and full text search built from first principles on object storage, 10x cheaper and extremely scalable. Uh, our next guest,