Research Engineer

About Decision-making and Reasoning Team

Decision-making & Reasoning Team has multiple divisions and is distributed across the globe. London division focuses on Reinforcement Learning, Bayesian Optimisation, and AI-based Agent Systems.

The international collective of scientists and research engineers covers multiple areas in Artificial Intelligence, Statistics, Optimisation, and Optimal Control.

The Team has demonstrated an excellent publication track in top Machine Learning and Computer Science journals (JMLR, MLJ, etc.) and conferences (ICML, NeurIPS, ICLR, DAC, etc.) and won multiple awards:

• NeurIPS 2020 Competition on Black-Box Optimisation,
• 2022 International Workshop on Logic Synthesis Open Competition,
• Best Paper Awards and Nominations on CoRL and IROS.

Job Summary

Large Language Models (LLMs) have shown remarkable abilities to perform reasoning in various tasks. By training on large natural language corpora, LLMs have developed a wide range of knowledge and linguistic abilities, allowing them to perform complex reasoning tasks. They can process and understand natural language, as well as generate coherent responses. Recent works have shown the LLMs' potential in autonomous agents. They can tackle complex decision-making scenarios conditioned on provided information and generate insightful recommendations. With these abilities, LLM-based AI Agents have immense potential to assist humans in several intellectual tasks and contribute to our understanding and solving of complex problems.

We are looking for Research Engineer/Scientists with prior experience in LLMs and AI Agent. In this role, the successful candidate will conduct academic and applied research in the field of AI Agent. Opportunities for collaboration with world-class academic organizations are provided. The long-term goal of this project is the utilization of LLM-based AI Agent for decision-making and reasoning. The candidate will work on individual or team projects which will involve the utilization of LLMs for the development of planning, reasoning, memory, reflection, tool-use and communication abilities.

Key Responsibilities:

· Carry out a literature review and investigate the state-of-the-art AI Agent frameworks and models.

· Work on solving important challenges and conduct cutting-edge research in LLM-based AI Agent.

· Build benchmarks/baselines using public datasets.

· Finetune LLMs on novel, applicable and practical tasks.

· Model evaluation and refinement.

· Publish research papers at top-tier NLP/ML/AI conferences and journals.

Successful candidates should be/have

· Published in top-tier NLP/AI/ML conferences and top-tier journals as the first author.

. Creative and innovative.

· Able to convert an idea into a prototype efficiently, optimize its cost and drive the implementation.

· Excellent knowledge of existing LLMs and AI Agent.

· Result-driven with good communication skills; be able to work efficiently in a multi-cultural, multi-site, multi-language and changing environment.

· Excellent root cause analysis and problem-solving skills.

· Focused, pro-active and motivated.

· Open-minded and able to learn fast.

· Able to work autonomously and as part of a team of research and technical experts.

· Broad knowledge, be versatile and hands-on.

· Demonstrated ability to generate new ideas and innovate.

Ideal candidates should have (desirable)

· A PhD degree in Natural Language Processing/Machine Learning or related areas, with a focus on LLMs and AI Agent.

· A strong research track record, and have published in top-tier NLP/AI/ML conferences including NeurIPS, ICLR, ICML, ACL, NAACL, EMNLP, EACL, and top-tier journals.

. Excellent knowledge of LangChain, Auto-GPT, RL, MAS.

If you wish to apply for this position and believe your experience and skills are suitable, please email a copy of your CV and cover letter to