Patronus AI develops simulation research and infrastructure aimed at accelerating the progress towards human-aligned Artificial General Intelligence (AGI). The company provides evaluation models and tools designed to detect issues such as hallucinations, safety risks, and reasoning failures, as well as support custom evaluation workflows. Patronus AI also offers datasets, benchmarking capabilities, and automated evaluators for text, image, and multimodal AI applications. In addition, it conducts research on digital world models and simulation environments intended to support the training and assessment of increasingly capable AI agents.
On-site & Remote