Research Scientist / Founding Member – Agentic AI

Research Scientist / Founding Member – Agentic AI

? Location: Paris, France
? Focus Areas: Agentic Reinforcement Learning | Agent Memory | Self-Evolving Agents
? Team Structure: Early-stage, high-caliber research unit

About the Role

An emerging advanced AI research team based in Paris is building a small, elite group focused on foundational challenges in Agentic AI.

The mission is to push the frontier in:

• Agentic Reinforcement Learning

• Long-term and structured agent memory

• Self-evolving and self-improving agents

• Tool-using and multi-step reasoning systems

• Autonomous task decomposition and planning

This is a core research role focused on advancing next-generation agent architectures — not application-layer integration.

Selected candidates will join as founding research members and help define the long-term technical direction of the lab.

Position Overview

The team is seeking a Research Scientist / Core Technical Expert in Agentic AI with deep architectural understanding and hands-on system-building experience.

This role combines:

• Research originality

• System-level design thinking

• Engineering depth

• Practical evaluation of autonomous systems

The ideal candidate brings experience from leading AI research environments or equivalent frontier research settings.

Key Responsibilities

• Conduct cutting-edge research in Agentic AI, including:

  • Policy learning in tool-using environments

  • Long-horizon planning and credit assignment

  • Episodic, semantic, and structured memory architectures

  • Self-reflection and recursive self-improvement mechanisms

  • Continual learning in autonomous systems

• Design and implement new agent architectures beyond prompt engineering.

• Develop experimental frameworks for evaluating agent autonomy and performance.

• Contribute to high-impact research outputs (publications encouraged).

• Collaborate within a small, high-caliber research team.

• Help define the long-term research roadmap and technical agenda.

Required Qualifications

• PhD in Machine Learning, Artificial Intelligence, Reinforcement Learning, or related field.

• 2–6 years of research or industry experience.

• Deep expertise in at least one of the following:

  • Reinforcement Learning

  • LLM-based agents

  • Neural memory systems

  • Autonomous planning

  • Self-improving AI systems

• Strong system-building and implementation capabilities.

• Experience training or fine-tuning large-scale models.

• Excellent Python and ML engineering skills (e.g., PyTorch, JAX).

Preferred Qualifications

• Experience building real-world autonomous agent systems (tool use, long-horizon reasoning).

• Publications in top-tier ML conferences (e.g., NeurIPS, ICML, ICLR).

• Experience in advanced AI research labs.

• Familiarity with RLHF, RLAIF, or large-scale policy optimization.

• Knowledge of evaluation frameworks for agent performance and robustness.