Microsoft builds synthetic enterprise data to evaluate Copilot without customer data
AIMicrosoft Research says it built Industry Evaluation Data Generation Platform, which creates fully synthetic enterprise environments for testing Microsoft Copilot. The environments are grounded in public sources including O*NET occupational data, online job postings, and SEC filings, and automated content-safety checks screen the generated data for harmful or personally identifiable material.
Loading source posts…
Source details
Microsoft Research · Somya Chatterjee, Harshita Sahijwani, Min Chen, Sushmitha Rajeswari Muppa, Dayquan Julienne, Irene Shaffer, Chandra Inguva, Jenny Chen, Minha Hwang, Sam Ade Jacobs · microsoft.comPublished · added here
Original title: From synthetic enterprise data evaluations to real Microsoft Copilot AI confidence