We started with one obsessive question: can hiring be made deterministic? We've been building toward that answer since 2023, and we need the person who's going to build the engine that gets us there.
You'll sit with recruiters on Monday, figure out what's broken, and ship something into their hands by Thursday. No Jira tickets, no design specs handed to you — just a hard problem and full ownership.
You'll build and own the full agent stack — sourcing, screening, interview assist, fit evaluation — plus the eval harness that keeps it honest. When recruiters need a workflow surface to use what you built, you'll build that too, front to back. When a fine-tuned 3B model beats a frontier API call, you'll know and ship that instead.
2–4 years of engineering experience, AI-native by instinct — reaching for agent frameworks before being asked
Shipped a real agent system: tool use, memory, evals, and the unglamorous plumbing included
Built full-stack applications end to end and owned them in production
Hands-on with SLMs — LoRA, quantization, tokenizers, serving trade-offs, no bluffing
Can talk to recruiters like a human being — they're your closest collaborators
The determinism question makes you want to find out, not roll your eyes
Most AI jobs are "add a chatbot and call it AI-native." This isn't that. Your decisions become the architecture. The good ones and the embarrassing ones.
If that sounds fun, let's talk.
InCommon is an equal opportunity employer. We're building a diverse team because it makes us better at the problem we're trying to solve.