This position is ideally Paris based and French speaking but we will accept remote (EU based) candidates
🧠 What You’ll Do:
- Design and implement agentic AI architectures for dynamic task environments using frameworks like AutoGen, LangGraph, CrewAI, MetaGPT, or similar.
- Integrate LLMs (GPT-4o, Claude, Gemini, Mistral, etc.) into multi-agent systems with persistent memory, tools, and planning capabilities.
- Develop and fine-tune reasoning agents that interact with APIs, databases, vector stores, and real-world systems.
- Architect tool-using agents that can browse, retrieve, write code, summarize, and collaborate.
- Collaborate with frontend/backend developers and product designers to ship prototypes and demos rapidly.
- Contribute to system design, scalability, observability, and prompt strategies across complex agent workflows.
- Stay current with the latest research in autonomous agents and operationalize novel methods.
- Core Languages: Python (FastAPI, asyncio), TypeScript (for UI/agent interfaces)
- LLMs & Providers: OpenAI (GPT-4/4o), Anthropic, Mistral, Cohere, Ollama
- Agent Frameworks: LangChain, AutoGen, CrewAI, Semantic Kernel, LangGraph, Haystack, Guidance
- Memory & Storage: Chroma, Weaviate, Redis, Pinecone, Postgres, Milvus
- Execution & Orchestration: Docker, Ray, Celery, Dask
- Tooling: OpenAI Tools / Functions, LangChain agents, Retrieval-Augmented Generation (RAG), Vector Search
- DevOps (Optional): Kubernetes, CI/CD, observability stack