Lead Research Scientist in LLM Development
Join an innovative research group dedicated to pioneering AI systems that excel in reasoning, planning, and acting across diverse physical environments. Our mission is to develop intelligent agents that can experiment, create, and engineer solutions that significantly boost scientific and industrial advancement.
This dynamic team merges a wealth of technical expertise with a track record of practical achievements, including substantial government-funded projects. We thrive at the intersection of advanced model research and robotics, simulations, and automated engineering systems, delivering impactful solutions born from rigorous scientific principles and bold execution.
Why You'll Find This Role Exciting
• Engage in groundbreaking research on reasoning, planning, and tool-use models that directly govern autonomous engineering systems.
• Expand the frontiers of SFT, RLHF, DPO, verifier-guided RL, and long-horizon planning, with your research translating into immediate real-world applications.
• Be a part of a fast-paced research culture alongside exceptional colleagues focused on agent systems, simulations, data management, and intricate toolchains.
• Enjoy significant ownership within a small team tackling one of the most challenging technical endeavors of our time.
Role Overview
We are seeking a Research Scientist specializing in LLMs to drive the development of next-generation reasoning and agent architectures. Your contributions will encompass model innovation, alignment strategies, structured tool orchestration, and experimental applications of agents in real engineering settings.
This position offers a unique blend of deep research engagement and hands-on systems integration, providing the autonomy and scope necessary to lead foundational advancements.
Key Responsibilities
• Design and develop cutting-edge models and prompting systems for planning, multi-step reasoning, and structured tool utilization.
• Lead training efforts focused on SFT, RLHF/DPO, verifier-guided RL, and modular expert frameworks to enhance robustness and controllability.
• Establish schemas, tool-calling tactics, policy constraints, safety protocols, and recovery mechanisms for agent operations.
• Collaborate extensively with engineering, simulation, and data teams to implement, train, and assess models integrated within real-world production-like frameworks.
Qualifications
• Extensive experience in LLM research, agent reasoning models, or structured tool-use methodologies.
• Strong foundation in working with SFT, RLHF, DPO, or reinforcement-learning-from-verification techniques.
• Proven ability to design, analyze, and enhance long-horizon behaviors and decomposition methodologies.
• Comfortable navigating ML research, systems engineering, and hands-on experimentation in a dynamic environment.
• A history of excellence and ownership in technically rigorous fields.
