Alternance - Ingénieur.e IA Systèmes Agentiques MultiLLM H/F

About Thales

Thales is a global leader in high technology, specializing in three business sectors: Defense & Security, Aeronautics & Space, and Cyber & Digital. It develops products and solutions that contribute to a safer, more environmentally friendly, and more inclusive world. The Group invests nearly €4.5 billion annually in Research & Development, particularly in key innovation areas such as AI, cybersecurity, quantum, cloud technologies, and 6G. Thales has nearly 85,000 employees in 65 countries.

Our Commitments, Your Benefits

• Our technological expertise
• Our focus on employee well-being
• An inclusive and supportive environment
• Recognized social and environmental commitment (Thales Solidarity, CAC 40 ESG index...)

Your Daily Life

Join us at the Vélizy campus, which hosts a wide diversity of Thales activities, including aerospace and defense. Located west of Paris, this site offers a work environment equipped with the latest technological innovations and provides services to help balance professional and personal life.

Within Thales's PRS department, you will join an innovative team working on the development of multi-LLM agentic systems deployed in an on-premise environment.

You will contribute to the design and industrialization of advanced artificial intelligence solutions, integrating intelligent agents capable of orchestrating language models, processing complex data, and addressing high-value business use cases. You will work in a rich technical environment, at the intersection of AI, distributed architectures, and data systems.

You will be involved in the following four areas:

1. AI Agent Development

• Implement and improve specialized agents
• Design new nodes and edges in the multi-agent orchestration graph
• Write and optimize system prompts, Ollama Modelfiles, and agent configurations
• Develop agentic skills
• Set up unit and integration tests for agents and skills

2. Vector Memory and Continuous Learning

• Implement the embedding and continuous learning pipeline
• Evaluate embedding quality and improve semantic search relevance

3. MCP Connectors and System Integrations

• Develop and maintain MCP servers (FastMCP)
• Implement new MCP tools based on business needs
• Test, document, and monitor each connector with real use cases

4. LLM Evaluation and Quality

• Design and execute benchmarks to compare on-prem model performance
• Set up automatic evaluation pipelines (LLM-as-judge, RAGAS, custom metrics)
• Contribute to fine-tuning efforts
• Measure quality metrics: confidence, relevance, coherence, response time

Your Profile

You are pursuing a Bac +4/5 degree (engineering school or university) specializing in AI, Machine Learning, Data Science, or distributed systems, and you are looking for an apprenticeship?

You have:

• Advanced proficiency in Python
• Good command of LLMs and prompt engineering
• Experience with frameworks like LangChain or LangGraph
• Knowledge of vector databases: Qdrant, FAISS, Chroma
• Experience in REST API development (FastAPI)
• Familiarity with Docker and GitLab CI/CD pipelines
• Understanding of RAG approaches and embeddings
• Knowledge of SQL / PostgreSQL
• Proficiency with Git / GitLab tools
• Experience with Linux / Bash environments
• Ideally, familiarity with MCP or agentic frameworks
• Ideally, knowledge of LLM evaluation methods (RAGAS, etc.)

Are you recognized for your autonomy, analytical skills, curiosity, teamwork, and proactivity? Then this position is for you!

A Word from the Team

By joining our team, you will work at the forefront of the latest artificial intelligence technologies, with a focus on agentic systems and LLMs. You will be mentored by passionate experts and contribute to innovative projects with significant impact. This apprenticeship will allow you to develop recognized expertise in rapidly evolving fields and become a key player in the intelligent systems of tomorrow.

Innovation, passion, ambition: join Thales and create the world of tomorrow, today.

The position may require access to information classified as national defense secret. The selected candidate will undergo a security clearance procedure in accordance with articles R.2311-1 et seq. of the French Defense Code and the IGI 1300 SGDSN/PSE of August 9, 2021. Thales, a company committed to disability inclusion, recognizes all talents. Diversity is our greatest asset. Apply and join us!