Before starting her own company, Kalyani Ramadurgam worked at Apple, where her job was to make sure people on terrorism watch lists couldn’t use Apple Pay. It was serious, high-stakes work. But what surprised her most was how it was done. Even at one of the world’s most advanced tech companies, compliance—the process of following financial laws and regulations—was still slow, manual, and painfully outdated. “It meant reading through not just hundreds, but thousands of pages of documentation,” she says according to Forbes. Teams were overwhelmed. “Organizations were just throwing bodies at the problem.”That experience stuck with her. She earned a spot on Forbes 30 Under 30 USA 2026 in Finance. In 2023, Ramadurgam co-founded Kobalt Labs with Ashi Agrawal, a former software engineer at Affirm, to bring compliance into the machine-learning age. Their idea was simple but ambitious: use AI to do the heavy lifting humans were drowning in.Kobalt’s AI models sift through massive volumes of documents to help banks and financial institutions vet their business partners. The system checks whether companies are following critical rules—such as stopping money transfers from sanctioned countries or quickly disclosing security breaches—tasks that once took teams weeks or months to complete by hand. The approach is already gaining traction. Kobalt Labs has raised $13 million in funding and now works with more than 20 customers, including Celtic Bank and fast-growing fintech company Bilt.For Ramadurgam, Kobalt is about more than efficiency. It’s about modernizing a system that protects the financial world—without burning out the people responsible for keeping it compliant.
