NRI PULSE STAFF REPORT
Seattle, WA, Oct. 21, 2025: Navigating America’s complex and often frustrating employment-based visa system could soon become easier, thanks to an Indian techie who turned her own struggles into a startup solution.
Priyanka Kulkarni, a 34-year-old machine learning scientist who spent nine years on various U.S. visas, has launched Casium, a Seattle-based artificial intelligence platform designed to simplify and streamline the immigration process for skilled workers and employers alike.
Kulkarni, who previously worked at Microsoft, says her experience with the U.S. immigration system — marked by uncertainty, delays, and complicated paperwork — inspired her to build a better way. “I knew there had to be a smarter, more transparent approach to help people like me,” she said in an interview.
Casium, founded in 2024 and backed by $5 million in seed funding, uses AI to guide users through every stage of employment-based immigration — from assessing eligibility and recommending visa types such as H-1B, O-1, or EB-1A, to automating document preparation and coordinating with licensed immigration attorneys. The platform promises to cut preparation time for certain cases from several months to less than 10 business days.
The U.S. work visa process is notoriously complex, involving multiple categories, strict compliance requirements, and shifting policies. For employers, the burden of hiring global talent can include significant legal costs and risks. For foreign professionals, the stakes are even higher, with visa backlogs, unpredictable timelines, and potential career disruptions.
By applying machine learning to this maze, Casium aims to bring efficiency and clarity to a system that has long been criticized as outdated. Experts say tools like this could help companies reduce costs and compliance errors while empowering skilled immigrants with better information and faster filing options.
Cover photo credit: schedule.sxsw.com.
