Scale AI

Scale AI is a San Francisco company founded in 2016 by Alexandr Wang that builds the data infrastructure many AI teams use to train and evaluate their models. Through its Data Engine, it sources and labels training data for use cases including generative AI, autonomous vehicles, robotics, AR/VR, and mapping, relying on a human-in-the-loop annotation workforce run through its Remotasks and Outlier subsidiaries. The company states that roughly a quarter of its contributors hold advanced degrees, and that most of the industry's leading generative AI model builders use its data pipeline. Its product line has grown beyond labeling into deployment and evaluation: the Scale GenAI Platform helps enterprises put AI systems into production with human oversight, Scale Donovan targets applied AI use cases, and Scale's evaluation tools include private leaderboards for benchmarking model performance. Reported customers include OpenAI, Meta, and the US Department of Defense, with a dedicated Scale Federal unit serving government and defense work alongside clients in automotive, healthcare, and other regulated industries. Meta has invested $14.3 billion in Scale for a 49% ownership stake. Since so much of the pipeline runs through a large outsourced annotation workforce rather than fully automated tooling, the consistency of results can depend on that labor pool.