Research Scientist/Engineer - General Decision & Control Agent

Research Scientist/Engineer - Agent Systems & Reinforcement Learning

Location: London
Salary: £(phone number removed) per annum + permanent benefits + bonus
Job Type: Permanent, Full-Time, On-site

About the Opportunity

We are partnering with a leading AI research organisation focused on developing sustainable, generalisable and evolvable Agent systems that represent the next frontier of artificial intelligence. This team is exploring how autonomous AI agents can operate effectively across complex environments, continuously learn from experience, and improve their capabilities over extended periods of execution.

This is an exceptional opportunity to join a world-class research environment working at the intersection of Agents, Large Language Models, Reinforcement Learning and Autonomous Systems, contributing to cutting-edge research that could play a significant role in advancing the path towards Artificial General Intelligence (AGI).

You will work alongside internationally recognised researchers and engineers, with access to significant computational resources and the freedom to investigate ambitious research challenges while helping translate breakthrough ideas into practical AI systems.

The Role

As a key member of the research team, you will contribute to the design and development of next-generation Agent systems, focusing on long-term reasoning, memory, self-improvement and reinforcement learning-driven optimisation.

Key Responsibilities

Agent Memory & Long-Term Reasoning

Design and develop advanced Agent memory architectures capable of supporting ultra-long context processing.
Research techniques to mitigate memory degradation in long-running Agent environments.
Improve information retrieval, storage and utilisation mechanisms to enhance long-term planning and decision-making.
Explore scalable approaches for persistent memory systems across complex task environments.Agent Self-Evolution & Autonomous Learning

Develop self-evolving Agent capabilities that enable continuous improvement through experience.
Research unified Agent representations and optimisation frameworks to support autonomous adaptation.
Build systems that facilitate iterative self-improvement and long-term learning.
Contribute to the development of Agent Harness frameworks that enable scalable evolution of autonomous agents.Agentic Reinforcement Learning

Investigate advanced reinforcement learning methodologies for Agent optimisation.
Develop both parametric and non-parametric RL approaches to improve Agent performance.
Build collaborative update pipelines connecting Agent policy models and Agent execution frameworks.
Evaluate and improve learning efficiency across diverse environments and task domains.Research & Innovation

Conduct novel research in Agent systems, Large Language Model reasoning and reinforcement learning.
Publish findings and contribute to the broader AI research community.
Collaborate with multidisciplinary teams to translate research breakthroughs into practical systems.
Stay at the forefront of emerging developments within autonomous AI and intelligent agent technologies.About You

Essential Requirements

Bachelor's degree or higher in Computer Science, Artificial Intelligence, Mathematics, Statistics or a related technical discipline.
Strong engineering implementation skills and/or deep theoretical foundations in machine learning and AI.
Demonstrated expertise in at least one of the following areas: Agent Systems, Reinforcement Learning, Large Language Models, Autonomous AI or Reasoning Systems.
Excellent programming skills, particularly in Python and modern machine learning frameworks.
Strong analytical and problem-solving abilities with a passion for tackling complex research challenges.
Excellent communication and collaboration skills within multidisciplinary research teams.
Commitment to working within a highly ambitious and fast-paced research environment.
First-author publications at leading conferences including NeurIPS, ICML, ICLR, ACL, EMNLP or equivalent.Desirable Requirements

Experience conducting cutting-edge research in Agent systems, reinforcement learning or foundation models.
High-impact research contributions, highly cited publications or influential open-source projects.
Track record of success in relevant competitions such as Kaggle, ARC-AGI or similar AI benchmarks.
Experience developing large-scale AI systems within industry, academia or advanced R&D environments
Only candidates based in UK and eligible to work in UK are allowed