About FlowMarket
FlowMarket is an AI agent network for B2B sourcing. It matches industrial buyers with qualified suppliers, using both demand buyers post and demand we detect for them. We connect existing industry directories and platforms instead of competing with them.
The role
You build and run the core platform: the matching engine, the agent layer and the directory integrations. You report directly to the CEO.
Responsibilities
- Develop and improve the matching pipeline (demand inference, profile filtering, LLM-based quality scoring, outreach limits)
- Build and maintain agent-to-agent conversations and RFQ flows
- Integrate directories and partner platforms (APIs, scraping, data pipelines)
- Evaluate and tune LLM usage for quality, cost and latency across providers
- Own infrastructure, monitoring and data quality
Stack
React / TypeScript, Supabase (Postgres, Auth, Edge Functions), Vercel, OpenAI / Anthropic / DeepSeek APIs
Requirements
- 4+ years of full-stack development, TypeScript and Postgres in production
- Hands-on experience building LLM features: prompting, evals, structured outputs, agent workflows
- Experience with data pipelines, scraping or third-party API integrations
- Pragmatic: you ship, measure and iterate
- Good English; German is a plus
- Plus: embeddings / vector search (pgvector), matching or recommendation systems, MCP / A2A, marketplaces
We offer
- Salary: competitive + bonus
- Remote work
- Short decision paths, direct work with the CEO
Apply
CV or GitHub, plus 3–5 sentences on how you'd score match quality between a buyer's need and a supplier profile.