Senior AI Engineer – LLM / Agentic Systems (f/m/x)

Technologies we use

About the project

This is how we organize our work

This is how we work

Your responsibilities

• Enable go-live and continuous improvement of the CES AI Assistant — a natural-language interface over the CES CubeJS analytics and semantic layer (90+ cubes) built on AWS Bedrock
• Design and maintain the agent's tooling (MCP servers, domain-aware tools) so the LLM reliably grounds answers in a large structured schema and returns correct results
• Build and run the evaluation pipeline that keeps a non-deterministic system trustworthy in a healthcare context, with per-user row-level security and no exposure of sensitive data

Our requirements

• Python and TypeScript/JavaScript
• LLM application development (prompt design, tool/function calling, MCP, RAG, agentic workflows)
• One or more of: Strands, LangChain, Pydantic
• AWS API Gateway & Lambdas (Node.js / Python)
• Terraform / Infrastructure as Code
• Docker, Docker Compose
• Clean code principles and evaluation-driven development

Optional

• AWS Bedrock (or equivalent managed LLM platform)
• Broader AWS stack (Cognito, ECS Fargate, S3, CloudFront, IAM)
• GitLab/GitHub
• AWS Redshift and CubeJS
• Semantic Layer
• Database management (SQL, noSQL)

What we offer

• Flexible engagement — preferably 100% capacity, but availability can be discussed
• Remote collaboration with a team working primarily in the Central European Time (CET) time zone
• Medicover
• Medicover Sport
• 100% remote
• Start ASAP

Benefits