### Join Substrate [Substrate](https://substrate.run/) is an expressive graph SDK, a pristine library of nodes distilling the latest AI models and techniques, and state-of-the-art infrastructure for serving multi-inference workloads. Companies like [Substack](https://substack.com/), [Maven](https://maven.com/), and others use Substrate to quickly protoype and deploy workflows combining dozens of inference modules. We're obsessed with **craft & beauty in developer tools**, and we're on a mission to bring a Stripe-quality API experience to the world of AI engineering. We do many of the things which under the hood make Stripe a world-class developer experience: versioning machinery to ensure extensive backwards-compatibility... a code generation system that produces SDKs, API documentation, and runnable demos... But all of this is instrumental to the bigger vision, which is tackling a big infrastructure problem. It's too hard to build complex multi-model systems. And we believe the solution requires much more than just chaining API requests – it should be possible to run dozens of models simultaneously, orchestrated in a larger logical system. Some large orgs have the compute, infra, and internal SDKs to do this effectively, but it's out of reach for most developers. ### Roles - We're looking for engineers with good taste and flexible skills. Technologies we use include Pulumi, Ray, and Jinja. - We're looking for the right community-oriented engineer to lead developer relations. ### Team We mostly work in-person from our office in East Williamsburg. Sometimes we have offsites, like this one in Austin. ![team-loro](https://media.substrate.run/team-loro.png) ### Apply To apply, email ben `@` substrate.run