Description
Summary We're looking for an experienced AI prompt engineer(preferably French speaking) to take over and continue developing a structured knowledge base that powers a conversational AI assistant deployed in a contact center environment. The system is already in active production with a mature versioning discipline: the knowledge base is decoupled from prompt logic, changes are tracked across versions, and deployments go through a tested pipeline before release. You'll work alongside our existing tester to maintain and extend the knowledge base, refine prompts based on real usage feedback, and continue migrating model invocation from direct API calls to AWS Bedrock. A web-based test runner (SSE-based, parallel execution, run history, Excel export) is already in place for validating changes before deployment - you'll use and extend it as the knowledge base grows. Scope of work: -Maintain and structure the knowledge base (versioned markdown formats) -Write and iterate on prompts -Run regression tests using the existing tool -Document changes for knowledge transfer What we're looking for: -Real prompt engineering experience — not just writing standalone prompts, but structuring versioned knowledge -Python proficiency -Familiarity with AWS (Bedrock) -Methodological rigor around testing and versioning -Ability to document work for smooth knowledge transfer -Prior experience with contact center or customer service systems is a plus, not required