evaluation
(5)
Instance
Instance is a San Francisco startup building the autonomous evaluation rig for robot learning. Its first product is a success detector that watches a robot's rollout video, reads the task description, and returns a verdict on whether the task was completed - with a subtask-by-subtask breakdown. Founded in 2026 by MIT classmates and childhood friends Claire Mao and Lucy Cai, Instance is part of Y Combinator's Summer 2026 batch and aims to remove humans from the tedious loop of watching robots, grading them, and resetting the scene.
Roark
Roark is an observability, testing, and quality-assurance platform for voice AI agents. It lets teams replay real production calls against new agent versions, catch failures before they ship, track caller sentiment and vocal cues, and monitor live voice interactions with 40+ built-in metrics. Founded in 2025 and backed by Y Combinator (W25), Roark says it processed more than 10 million minutes of calls in its first six months.
Distributional
Distributional is an enterprise AI testing platform that uses adaptive, statistical analysis of production traces to catch behavioral drift and regressions in AI agents and LLM applications. Founded in 2023 by former Intel AI software GM Scott Clark, the company helps Fortune 500 teams ship generative AI with confidence by treating reliability as a continuous test problem rather than a one-shot benchmark.
Deccan AI
Deccan AI is a Mountain View-based GenAI data company that runs the post-training and production layer for frontier labs and enterprises. It builds super-accurate SFT and RLHF datasets, reinforcement learning environments and agentic evaluation pipelines using an elite expert network and a purpose-built quality platform.
Scale AI
Scale AI is a San Francisco-based AI infrastructure company founded in 2016 by Alexandr Wang and Lucy Guo. It provides the data engine, evaluation tools, and AI deployment platforms that power the world's leading AI labs, Fortune 500 enterprises, and US government agencies. By combining a massive distributed workforce with proprietary tooling, Scale accelerates AI development through high-quality data labeling, RLHF, model evaluation, and agentic platforms — making it one of the most consequential picks-and-shovels companies in the modern AI boom, with a $29B valuation as of mid-2025.