Benchify is building the reliability layer for AI-generated code. Today, we make sure dynamically generated code always builds. Tomorrow, we’ll add correctness and security guardrails so any code written by an LLM just works. Our vision is instant, self-healing code pipelines — where AI-generated software not only compiles, but runs reliably and securely without human babysitting.
Requirements
- Have strong experience with containers, runtimes, and DevOps tooling (Docker, K8s, , OCI images, AWS, etc.).
- Are fluent in TypeScript, Node.js, or Python.
- Especially if you know a decent amount about the underpinnings of these languages – e.g., if you can explain why uv is faster than pip, if you have strong opinions on Parcel versus Vite, or if you feel that Python “doesn’t really have a type system”.
- Care about performance and correctness, not just “works on my laptop.”
- Are curious about compilers, testing, or program analysis (bonus points).
Responsibilities
- Design and optimize containerized environments that can boot, run, and heal code in milliseconds.
- Build infra that validates, repairs, and runs AI-generated code at scale.
- Work across the stack when needed — but especially on the dev/ops side of things.
- Talk to customers and ship features that solve real-world problems.
- Help shape our engineering culture (lots of esoteric memes).
Other
- US citizen/visa only
- Any (new grads ok)
- In person in San Francisco or New York City
- Competitive salary + meaningful equity
- ASAP
