Forest AI | Remote (Europe, CET overlap) | ~10% travel |
## Start: ASAP
Full time preferred, four days works, three is possible. Freelance, permanent or B2B: your choice (freelance day rate discussed separately).
The best developers are now part product manager and part project manager of their own army of AI coding agents. Their job is making sure those agents deliver what the business actually needs, not just what someone wrote in a ticket. In that world the bottleneck isn't producing correct code. It's understanding the business and communicating clearly.
We work Claude Code first. If you don't believe in delivering that way, this isn't the right fit.
## About Forest AI
Forest AI helps companies become benchmark-driven in how they build AI agents. AI agent development demands an order of magnitude more testing before release than traditional deterministic software. We build the automated evaluation and testing layer that makes testing at that volume practical, so our customers can put AI into production in places where being wrong is expensive.
## What you'll actually work on
Your first engagement is with a German insurance venture (name shared in conversation) building an AI-driven B2C insurance value chain, quote, consult and purchase, all AI-driven, going live in September. There is a second track automating their internal insurance operations. It runs like a startup: decisions in-house, no corporate approval loops, ideas become working software the same day.
## The role
100% hands-on. You take a vague business need, turn it into a system that works, and keep it working, owning the why, the build and the test without depending on someone else for the biggest chunk of it.
- Pick up signal fast. Hear one sentence mid-meeting, recognise it carries real information about what the business needs, work out what the speaker actually wants, and go deliver it.
- Own outcomes to completion. We drop you into a project, say "automate this manual customer workflow," and three months later it just works, without anyone directing every detail.
- Own customer satisfaction. If the customer doesn't see the ROI, the software gets thrown out. Everyone here writes down why a feature matters to the customer. Sometimes that lands in a presentation. When it doesn't, it lands in the commit message.
- Push the business for clarity. Press for decisions where the right AI behaviour can't be intuited, or where the data to decide it doesn't exist yet.
## How we build
TypeScript end to end. Next.js and React on the front, Postgres on Supabase, AWS underneath, an agent framework (we use Mastra, any comparable one is fine), and Langfuse for tracing and evaluation.
Claude Code is the primary build tool, not an accessory. That's the part we care about most.
The specific tools matter far less than how quickly you become productive with them. If you've shipped serious software in another stack and you steer AI agents well, you're in scope.
## What we're looking for
- Several years building and running production software real customers depend on: APIs, backend services, data pipelines, and the operational reality of keeping them up.
- A track record of owning a customer-facing build from zero to production, ideally more than once.
- AI-native, hard requirement. Claude Code or Codex as your daily driver.
- Roughly a year of real, hands-on exposure to building LLM-based features in production. Nobody has five years of it. Claiming so is a red flag.
- Strong discipline around automated testing, CI/CD and monitoring. Evaluation and testing are our product, so this matters more here than in most places.
- Real customer-facing time: consulting, forward deployed engineering, founding your own company, or an in-house role where you sat with the business. You've run requirements conversations and can separate the requests that genuinely drive return from the wishes and curiosities that arrive alongside them.
- Pragmatic, solut
