Read the deck an ex-Waymo engineer used to raise $3.75 million from Sheryl Sandberg and Kindred Ventures

Scorecard, a burgeoning startup based in San Francisco, has successfully raised $3.75 million to develop an innovative platform for evaluating artificial intelligence systems. This platform enables customers to rigorously test their AI agents against established metrics, facilitating quicker deployment. The evaluation process assesses various aspects of an AI system, including performance, efficiency, and safety, in comparison to industry benchmarks. The concept was born when founder Darius Emrani was working on a healthcare application designed to transcribe patient notes for improved doctor-patient interactions. Emrani quickly recognized the potential dangers of AI errors, stating, "If it messed up, then it could hurt or kill people. I realized there's a need for evaluations." Emrani's background includes a tenure at Waymo, where he managed testing and simulations for self-driving vehicles for four years before departing in 2022. He noted that as numerous companies rush to launch AI applications, sectors like healthcare, finance, and legal technology face intense pressure to deliver swiftly. "There’s often an 'Oh crap' moment when they realize it doesn’t work," he explained, highlighting that some legal startups have faced significant failures, leading to severe consequences for lawyers involved. To address these challenges, companies turn to Scorecard's platform to assess the effectiveness of their AI agents and make necessary improvements. Emerging from stealth mode, Scorecard has reported conducting millions of tests for clients, including Thomson Reuters, which utilizes the platform to evaluate its suite of legal AI agents, CoCounsel. The funding round was primarily led by Kindred Ventures, with additional contributions from Neo, Inception Studio, and Tekton Ventures. Notable angel investors include former Meta COO Sheryl Sandberg, along with individuals affiliated with OpenAI and Apple. With this financial boost, Emrani, alongside a team of three engineers, intends to expand by hiring additional engineering and design talent, as well as a go-to-market specialist. "I envision a world where agents are universally accessible in essential services that significantly impact people's lives," Emrani expressed. He is enthusiastic about the rapid evolution of the AI landscape, noting, "My customers didn’t exist a year and a half ago. This whole space is so new."