Founding Applied Research Scientist, AI & Simulation

Summary:Early-stage investment is looking to hire a Founding Applied Research Scientist for a small technical team with proven founders (successful industry veterans). You’ll work closely with the CTO to develop and test machine learning methods, investigate where they fail, and help turn effective approaches into dependable product capabilities. This role combines independent research with hands-on implementation.Research Backgrounds of InterestWe’re interested in exceptional researchers from one or more of the following areas:Language models: training, post-training, fine-tuning, alignment, or personalizationReinforcement and preference learning: reinforcement learning, reward modeling, preference optimization, or learning from feedbackSimulation and agents: world models, generative agents, agent-based modeling, or multi-agent systemsDigital twins and physical systems: physical simulation, learned dynamics, surrogate models, system identification, or sim-to-real methodsHuman and social systems: computational social science, behavioral modeling, decision-making, or social simulationModel reliability: evaluation, calibration, uncertainty estimation, robustness, or learning from limited and noisy dataYou do not need experience in every area. We’re looking for strong ML fundamentals, sound experimental judgment, and the ability to build and test your own ideas. A PhD is welcome but not required. Direct work with frontier language models is especially valuable.You’ll work closely with the founders and engineers to frame research problems, compare approaches, analyze results, and bring successful methods into production.About GreylockGreylock is an early-stage investor in companies including Airbnb, LinkedIn, Dropbox, Workday, Roblox, and Coinbase. Our talent team helps candidates connect with opportunities across our network (no fees).This role is for direct employment with our investment.