Variance builds AI agents that automate risk and compliance workflows—including identity verification, KYC, fraud investigation, and compliance review—for large financial services firms and B2C companies. The platform replaces manual operations teams with autonomous agents, freeing up significant human effort at scale.

Michael Lin and Karine Mellata both have direct experience working on Trust & Safety teams at Apple and Discord, giving them unique insights into the problems Variance aims to solve. Trust & Safety is a significant and increasing challenge for companies with large user bases due to rising UGC and content creation, the spread of hateful content, and increasing regulation. Google and Facebook alone spend ~$2B annually on Trust & Safety, indicating a substantial market opportunity for a more efficient solution. Variance differentiates itself by focusing on owning the Trust & Safety data pipeline, similar to a Customer Data Platform (CDP). This approach is presented as more mission-critical and sticky than being a pure-play classifier or orchestration platform, offering data defensibility and a holistic solution that the founders believe is missing in the current market.