Mercor

Patronus AI provides frontier AI labs with training data, tasks, and simulated environments for reinforcement learning. The company builds RL environments that replicate real-world digital workflows across software engineering, computer use, customer support, and financial services. Each environment pairs simulated applications with tasks and reward signals, so models can practice long-horizon, multi-step work under realistic conditions such as latency, interface changes, and unexpected pop-ups. Patronus AI also develops Digital World Models, AI models trained to predict how digital environments behave, which are used to generate and evolve training environments at scale as models improve. Its work spans reinforcement learning environments and digital world models for agents operating in software, supporting post-training, benchmarking, and evaluation of frontier AI models and agents.

Foundire is an AI-powered hiring platform that connects every step of recruiting into one seamless workflow, letting teams search a global database of over 800 million professionals, automatically screen and rank resumes, run adaptive AI first-round interviews, analyze detailed insights and reports, and verify candidate backgrounds all in one place. It visually links job creation, candidate sourcing, resume scoring, AI-assisted dialogue, video, or voice interviews, and structured evaluation into a single automated pipeline, reducing manual handoffs and speeding hiring decisions with data-backed assessments. Foundire’s semantic parsing extracts skills and experience for accurate shortlisting, its AI interviewer conducts real-time adaptive conversations with role-based prompts, and its multilingual support enables consistent evaluation across regions. Hiring teams gain intelligent interview records with playback and collaborative scoring, standardized job templates, & dashboards.