Pareto.AI
Pareto.AI builds a platform that converts expert human judgment into training signals for reinforcement learning, including curated datasets, evals, and RL environments for frontier AI models. It sells to AI labs and model developers that need reliable reward models and grading pipelines to train and assess large models. Privately held and based in San Francisco, it cites work with Anthropic and Google DeepMind on frontier-model training.
Open Positions at Pareto.AI
1 open position
Own the full lifecycle of reinforcement-learning environments used by frontier AI labs, from scoping and containerized builds through deployment and production health. This hands-on role combines production software engineering, customer-facing research collaboration, and platform automation.