Engineer - Rag & Agentic & Knowledge Graph H/F - Collective Works
Les missions du poste
AI Engineer - Python, Agentic AI & Knowledge Graphs
Context
Our client, a SaaS software company operating internationally, is looking for an AI Engineer to strengthen its AI Engineering team.
Reporting to the Head of Data & AI, you will help build AI features embedded in the platform, deployed in both SaaS and on-premise environments.
We are looking for a Python backend developer who has delivered AI agents in production, with expertise in ontologies and knowledge graphs - a core requirement for this role.
Responsibilities
Design and develop backend components for AI features running in production.
Build AI agents, including multi-agent systems, workflow orchestration and LLM-driven reasoning.
Develop MCP servers and custom tools connecting agents to applications, APIs and databases.
Design RAG and agentic architectures combining vector search, knowledge graphs and ontologies.
Model business domain knowledge to support agent capabilities and semantic reasoning.
Set up an AI gateway such as LiteLLM for model routing, governance and cost control.
Implement observability and evaluation for AI services.
Mentor junior developers and contribute to AI engineering best practices.
Must-have skills
Python / FastAPI: expert-level backend development, REST APIs, microservices and asynchronous programming.
Generative AI in production: significant, proven experience beyond POCs.
Agentic AI: LangChain, LangGraph or equivalent, and multi-agent systems.
MCP and A2A: Model Context Protocol and Agent-to-Agent protocols.
RAG: vector search, embeddings and context optimization.
Ontologies and knowledge graphs: business domain modeling and semantic reasoning.
Production engineering: OpenTelemetry, Docker, CI/CD, Git and testing with pytest.
Databases: SQL, particularly SQL Server and MySQL, and vector search using Elasticsearch or equivalent.
Nice to have
Neo4j or another graph database - strongly preferred.
LiteLLM or another LLM gateway.
LangSmith, Langfuse or Arize Phoenix.
Google Cloud and GKE.
Profile
Senior backend developer with practical expertise in agentic AI.
Autonomous, pragmatic and able to contribute quickly.
Strong communicator, comfortable working in English with international teams.
Able to mentor junior developers and share engineering best practices.
Practical information
Start date: ASAP.
Duration: 3 months, renewable.
Location: Paris, with 1/2 days on-site per week.
Languages: French and English.
