AI / LLM Engineering (Werkstudent)
AI Tools
Tech Stack
Agent Workflow
Building agentic AI systems that turn natural language questions into reliable analytics via agentic workflows with tool/function calling, planning, and self-checks.
About the Role
CognitX AI develops privacy-focused AI solutions for data analytics and process automation, enabling users to ask natural language questions about data and receive trustworthy answers, visualizations, and automated actions.
Key Responsibilities: Design and enhance agentic workflows including tool/function calling, planning, and self-checks for analytics and task automation. Create adapters/tools connecting LLMs to internal and external services. Develop backend features using FastAPI with Pydantic validation and testing. Build evaluation metrics measuring accuracy, latency, and cost efficiency. Optimize prompts and retrieval strategies for privacy, reliability, and performance. Deploy containerized solutions using Docker and Kubernetes with CI/observability support. Document architectural decisions and share team learnings.
Required: LLM/AI agent experience (function calling, RAG, MCP, agent frameworks) plus ML fundamentals. Production-grade Python proficiency (typing, testing, packaging). API/microservices familiarity (FastAPI preferred), Git, and Docker containerization. Strong problem-solving and proactive ownership mindset. Currently enrolled in Computer Science, Data Science, AI, or related field.
Nice-to-Have: RAG stacks, Redis, PostgreSQL. Kubernetes, GitLab CI, observability tools. React/TypeScript UI development. EU privacy/GDPR interest and enterprise AI safety. Self-hosted/open model experience. Knowledge graph concepts.
Benefits: Work on privacy-first AI with real enterprise deployments. Mentorship from senior AI experts. Flexible hours, remote-friendly environment. Competitive Werkstudent compensation.