Fabriq runs large open models across distributed machines, behind one OpenAI-compatible endpoint. It is a hard systems problem with a large upside, and we are keeping the team small and senior.
Small team, real ownership
You own systems end to end, from design to the pager.
Open by default
We build on open models and open protocols, and we prefer to work in the open.
Honest by default
We show real numbers, real limits, and real progress. No vanity metrics.
Backend Engineer
Remote · Full-time · Founding engineering
PythonRustCloudDistributed systems
We are hiring two backend engineers who know cloud infrastructure and backend systems. You will build the core of the network: the Python services that authenticate, route, and meter inference, and the engine layer that shards models across a fleet of heterogeneous machines.
What you will do
- Design and build the Python/FastAPI services that route inference across a distributed network: the gateway, control plane, scheduler, and metering.
- Go deep in the engine's networking layer (Rust, over zenoh) where latency and correctness both matter.
- Run it in the cloud: containers, queues, autoscaling, observability, and the boring reliability work that keeps a network up.
- Model the hard parts of a distributed system: node discovery, health, retries, backpressure, and failure that is the normal case, not the exception.
What we look for
- Strong Python and backend fundamentals; comfort dropping into Rust for the engine's networking crate, or the appetite to.
- Real cloud experience shipping and operating backend services in production.
- Comfort with distributed systems: concurrency, networking, and the tradeoffs between them.
- A bias for shipping, measuring, and iterating over polishing in private.
Do not see your exact role but think you should be here? Reach out anyway and tell us what you would build.