Principal Evangelist MongoDB - Document Model & Vector Search & RAG

The first time I ran MongoDB in production it was a sharded cluster spanning two data centers, holding up a billion-dollar e-commerce site. That was about a decade ago at Urban Outfitters, back when I was the guy carrying the pager for it. I recently joined MongoDB as a Principal Evangelist on the Builder Relations team. I spent six years being loud about data streaming, and I could keep that up for six years because the technology genuinely earned it. That's the only way this job works. You cannot fake enthusiasm in front of a room of engineers. They can smell it before you finish your first slide. So the question I asked myself was straightforward: what else do I believe in that much? The document model is one of those ideas that quietly changes how you think about building an application. Vector Search put the database exactly where AI needs it to be, which is close to the data rather than bolted on beside it. Someone I trust made the introduction, everything I had seen of Atlas from the partner side at Confluent backed it up, and the decision stopped being difficult. There's something else I'm looking forward to. A big part of the streaming job was explaining why real-time mattered before I could explain anything else. Every builder already knows they need a database. I get to skip straight to the interesting part, which is what you can actually build with it. Twenty years in the layers everyone else abstracts away — storage, network, cloud, streaming, data. Now helping builders get the data layer right so their AI actually works in production. Keep an eye out. I'm going deep on the document model, RAG, Vector Search, and plenty more, and I'll be spending much of my time putting other builders in front of a camera so they can show you what they've shipped. If you're building something interesting on MongoDB, I want to hear about it. That's the whole job.