Build the future of vector search

We're a small, focused team building a distributed vector database from scratch in Rust. Every engineer has direct impact on architecture, performance, and product direction.

How we work

Ship fast, measure everything

We deploy multiple times a day to production. Every change is measured against real query latency, throughput, and recall benchmarks. If it doesn't move a metric, we question whether it's worth doing.

Depth over breadth

We'd rather have one engineer who deeply understands SIMD-optimized distance calculations than three who can set up a Kubernetes cluster. We value expertise in systems programming, linear algebra, and distributed systems.

No complexity budget overruns

Every abstraction has a cost. We prefer three clear lines of code over a premature abstraction. Our codebase is small enough that every engineer can hold the full system in their head.

Our stack

Coordinator
Rust + Tokio
Workers
Rust + SIMD
Transport
WebSocket + bincode
Storage
Memory-mapped files
Client API
HTTP + JSON
Metadata
SQLite
Consumer Site
Node.js + EJS
Infrastructure
Bare metal Linux

Open positions

We don't have any open positions right now. We're a small team and hire opportunistically when we find the right people.

Interested anyway?

If you're passionate about systems programming, vector search, or distributed databases and want to work on hard problems with a small team, we'd like to hear from you.

Send us a note