About the role
Rally builds prompt-run B2B tooling for revenue teams. We're a distributed team of 9 (7 engineers) shipping to 400+ paying accounts. You'll join as our founding AI product engineer and own the model layer end to end.
Ship prompt-run B2B tooling with a small, senior team. Own the model layer end-to-end.
What you'll do
- Design and ship the prompt + eval pipeline (Claude, GPT-5, open models) that powers every product surface.
- Own retrieval quality: chunking, embeddings, hybrid search, and evals we can trust before shipping.
- Instrument every model call — cost, latency, quality — and turn the numbers into a weekly reliability dashboard.
- Partner with design to prototype net-new AI surfaces in Lovable, then harden the winners in production.
What we're looking for
- 3+ years shipping production software; at least 1 year shipping LLM-backed features to real users.
- Strong TypeScript + React. Comfortable in a Supabase / Postgres stack.
- You've written evals that caught a real regression (bonus: you can show them).
- Bias toward small, reversible releases and honest post-mortems.
Perks
Fully remote (US/EU overlap 4h/day)
Meaningful early equity
$3k/yr learning + conference budget
New laptop + $1.5k home-office setup
From the hiring managers
Direct"I want someone who ships weekly, writes evals before writing prompts, and treats model choice as a product decision — not a religion."
PS
Priya S.
Head of Engineering, Rally
"The best candidates I've seen came in through Build on Vibe — they arrive with a portfolio of shipped things, not just tutorials."
ML
Marcus L.
CTO, Rally
Apply for this role
Applications go straight to the hiring team. We aim to respond within 5 business days.