We're a small team doing work that would normally take a much bigger one — and the reason we can is that we automate aggressively. Research synthesis, reporting, parts of outreach: if it's repeatable, we want it automated, and that's what this role is responsible for.
You'll move fast between prototyping and hardening. Most ideas start as a rough script that saves someone an afternoon; the ones that prove out get turned into real tooling other people on the team depend on. You'll need judgment for which is which.
What you'll do
- Design and ship LLM-backed workflows into internal and client-facing tools
- Automate the repeatable parts of research, reporting, and outreach
- Prototype fast to test an idea, then harden what proves out into real tooling
- Work closely with the engineering team to integrate automations into existing products
- Keep an eye on cost, reliability, and failure modes as usage scales
What we're looking for
- Hands-on experience building with LLM APIs (OpenAI, Anthropic, or similar) in production, not just prototypes
- Solid scripting/engineering fundamentals — Python or TypeScript, comfortable gluing systems together
- A bias toward shipping something imperfect fast and iterating over it
- Judgment about when automation actually helps versus when it adds fragility
Nice to have
- Experience with agentic workflows or tool-use frameworks
- Familiarity with vector search / RAG pipelines
- A side project or two you can point to that automated something real