Location
Remote (UK), with a central London office 1 to 2 days a week for the first few weeks.
Salary
£90k–£100k + equity (sliding scale, trade cash for more equity if you want).
Product engineer at a seed-stage AI startup whose product lets people build and manage teams of autonomous AI agents.
The company
We’re working with a seed-stage, London-based startup building an AI agent platform. The product started life as an AI coding tool and has since turned into something stranger and more interesting: a place where people hire, instruct and manage teams of AI agents to do real work for them.
It’s growing on its own. Roughly a thousand new sign-ups a day, no paid acquisition. The model is freemium, and paid conversion climbs as people realise the agents are actually useful. They have around 18 months of runway and are aiming at a Series A in the next few months.
The team is tiny, around four engineers, remote-first. The bar is high and the ego is low. Nobody gets punished for breaking production. You’re expected to fix it and keep moving.
The role
You’d be a full-stack product engineer owning features end to end, including in production. The stack is TypeScript top to bottom: Bun, Fastify, PostgreSQL with Prisma, Redis, Next.js. The product leans hard on LLMs, with integrations across Anthropic, OpenAI, Gemini and OpenRouter, plus sandbox infrastructure for running agent code (E2B today, with a plan to build something proprietary on Firecracker once the Series A lands).
The interesting problems are real: streaming LLM conversations at scale, running untrusted agent code safely in sandboxes, and keeping model costs down without making the product worse. This is a product engineering role, not a pure backend one. They care more about whether the thing works for users than whether the code is perfect.
Honest version: the codebase is young and still being shaped, the product isn’t feature‑complete, and things move fast. If that sounds like chaos, it isn’t for you. If it sounds like ownership, it is.
What they’re looking for
- Strong TypeScript across the stack, around 3 to 6 years of experience
- High agency, the kind of engineer who owns their code through to production and watches the logs
- Product judgment, biased toward impact and revenue over perfect code
- Comfortable on Linux, happy doing their own deployment chores
- Has learned something from a production failure and can talk about it honestly
- Bonus: experience streaming LLM conversations, working in sandbox environments, or optimising model costs
Not a fit: purely backend or eg Python‑centric engineers. This is product‑facing TypeScript work.
- Remote‑first, occasional days in the central London office
- Trips to San Francisco (the CEO is based there). CTO is based in London.
The process
- Intro call with Ethiq (James), 30 mins
- Culture call with the tech lead, 30 mins
- Technical session with the principal engineer onsite. No whiteboard gotchas or on‑the‑spot puzzles. You show how you actually work day to day, using whatever tools you’d normally reach for (Claude, Cursor, whatever).
Two real stages after the intro, fast feedback, quick turnaround.
#J-18808-Ljbffr…