1 day ago
Luxembourg, LuxembourgMid Level
Responsibilities
- Build backend services and REST APIs in Python and FastAPI with asynchronous processing, validation, authentication, error handling, retries, and business-system integrations.
- Implement LangGraph and LangChain agentic workflows with state management, routing, tools, memory, persistence, recovery, and human-in-the-loop interactions.
- Integrate Azure OpenAI, OpenAI, Mistral, or IBM watsonx models while managing structured outputs, rate limits, latency, costs, and failures.
- Develop RAG ingestion and retrieval pipelines involving extraction, chunking, embeddings, vector databases, hybrid search, reranking, citations, and retrieval evaluation.
- Implement multi-agent coordination, MCP clients and servers, A2A integrations, shared context, task lifecycles, and failure handling.
- Evaluate and observe AI applications using automated tests, LangSmith or equivalent tracing tools, and response and retrieval quality assessments.
- Package applications with Docker, contribute to CI/CD, support cloud deployment, and apply authentication, access control, secrets management, and secure development practices.
- Use AI-assisted development tools for prototyping while reviewing, testing, securing, and maintaining generated code.
- Apply responsible AI practices related to privacy, transparency, traceability, human oversight, the EU AI Act, and GDPR.
- Document technical decisions and contribute to estimates, feasibility assessments, demonstrations, and collaboration with engineers, architects, business analysts, security specialists, and compliance specialists.
Requirements
- Strong practical Python and asynchronous programming experience, including async/await, asyncio, concurrent API calls, timeouts, retries, cancellation, and task management.
- Experience designing FastAPI and REST APIs with input validation, authentication, and consistent error handling.
- Hands-on experience with LangGraph, LangChain, LLM integration, RAG systems, multi-agent architecture, MCP, and A2A.
- Experience with Azure OpenAI, OpenAI, or equivalent model providers, including rate limits, latency, and cost management.
- Experience with document ingestion, chunking, embeddings, vector databases, hybrid search, reranking, citations, and retrieval evaluation.
- Experience with LangSmith or equivalent observability and evaluation tools for LLM applications and agent workflows.
- Experience with Git, Docker, CI/CD, pytest, unit testing, integration testing, asynchronous tests, and mocks.
- Experience deploying applications to a cloud platform, ideally Microsoft Azure.
- Familiarity with the EU AI Act and GDPR and their implications for privacy, transparency, traceability, human oversight, and responsible AI.
- Practical experience with Claude Code, GitHub Copilot, Codex, or similar AI-assisted development tools.
- Professional written and spoken proficiency in both French and English.
- Microsoft AI, Claude, cloud engineering, AI engineering, AI governance, or multi-agent architecture certifications are advantageous.
About Fujitsu
Fujitsu is a Japanese multinational IT services and computing company that provides systems integration, managed services, cloud, and enterprise hardware for businesses and governments. It also develops servers, storage, and client devices for enterprise workloads. Founded in 1935 and headquartered in Tokyo, Fujitsu is publicly listed on the Tokyo Stock Exchange and operates in more than 50 countries.