> Markdown version of [/jobs/ext/4501427-founding-engineer](https://www.wearedevelopers.com/jobs/ext/4501427-founding-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Founding Engineer

- **Company:** The French Sourcer
- **Location:** Berlin, Germany
- **Contract:** Permanent contract
- **Skills:** Application Programming Interfaces (APIs), Databases, Open Source Technology, Search Technologies, Retrieval-Augmented Generation, Large Language Models, Indexer, Agentic-AI, Data Pipelines, Serverless Computing
- **Published:** October 31, 2026
- **Apply:** https://www.adzuna.de/details/5946811048

## About the Role

* Solid full-stack experience with a track record of shipping scalable, production-grade systems
 * Hands-on experience with generative AI in production: RAG, vector search, LLM APIs or agent frameworks, beyond simple prompt integrations
 * Experience with search indexing and serverless databases
 * A builder profile, with projects you created from scratch, where side projects and open source count as much as job titles
 * Comfort making pragmatic decisions with incomplete information and iterating quickly based on real usage
 * Fluent English
 * Bonus: previous early-stage startup or founding team experience

## Description

Be one of the first engineers building AI agents that run recruiting end to end, at the stage where your architecture choices become the product.

What you'd actually work on

 * Owning the architecture and scalability of a platform already used by a growing base of customers, working directly with the founders and technical leadership
 * Building data pipelines for candidate information, from parsing CVs and documents to combining signals from several external platforms into one clean profile
 * Designing and improving the logic behind autonomous conversational agents that reach out to candidates, qualify them and schedule interviews
 * Building real-time matching between candidates and roles, and making it measurably better with each release
 * Taking features from idea to production yourself, shipping fast, reading how real users react, and iterating without waiting for a process to tell you what to do next

Where it gets technically interesting

 * LLM work that goes far beyond calling an API: retrieval-augmented generation, vector search and agent frameworks that must behave reliably with real candidates on the other end
 * Matching quality is the product, so every gain in ranking relevance has a direct, visible impact on customers
 * Messy real-world data: CVs in every format, incomplete profiles and conflicting signals from different sources that the pipeline has to reconcile
 * A modern full-stack setup with serverless databases and advanced search indexing, where latency and cost both matter as usage grows
 * A founding-stage codebase, so you shape the technical foundations instead of inheriting someone else's decisions