An ex-Microsoft scientist is building an AI startup to change how companies handle work visas

Priyanka Kulkarni, a 34-year-old machine learning expert with firsthand experience navigating the complexities of immigration, is on a mission to transform how businesses manage work visas. After spending nine years on a visa herself, Kulkarni has founded Casium, a startup that leverages artificial intelligence to streamline the employment-based immigration process. Casium offers employers a comprehensive portal to handle visa cases from start to finish, replacing the traditional reliance on cumbersome Excel spreadsheets and external law firms. This innovative solution emerges at a time of significant shifts in immigration policy, particularly following the Trump administration's unexpected executive order mandating a $100,000 fee for each new H-1B visa application. While some organizations embraced this change, many others were left scrambling and initiated legal action in response. With a focus on speeding up processing times and enhancing transparency, Casium has already aided hundreds of candidates through assessments, compliance reviews, and actual visa filings, boasting an impressive approval rate. Kulkarni notes that in many cases, founders utilizing Casium's services have moved from initial intake to having their employees on the job in less than a month. Recent funding rounds have further fueled the growth of Casium. Established in 2024, the startup secured $5 million in seed funding led by Maverick Ventures, with contributions from Ai2 Incubator, GTMfund, Success Venture Partners, and angel investor Jake Heller, who previously saw his startup acquired by Thomson Reuters. The process begins with candidates filling out an intake form, after which Casium employs a suite of software agents that autonomously gather relevant public data, including scholarly articles and patents, to generate a comprehensive dossier. Within minutes, the platform suggests the most appropriate visa options, such as H-1B, O-1, or EB-1A. The report is then sent to a network of independent, licensed attorneys and paralegals who work with Casium, simplifying the drafting of necessary legal documents. Kulkarni asserts that this technology significantly reduces the time required to compile application materials—from three to six months with traditional law firms down to under ten business days—and helps identify potential errors that could hinder the approval process. While Casium offers initial assessments at no cost, it charges a flat fee for visa filings, adjusted based on the type of visa and complexity of the case. The company is also exploring a subscription model for employers seeking ongoing support. The notion that software can facilitate navigation through the immigration system is gaining traction, with other startups like Parley, OpenLaw, Manifest Law, and Plymouth emerging to provide similar services. However, the transition from traditional legal practices to tech-driven solutions presents challenges, as employers must weigh the reliability of established attorneys against the capabilities of automation. Having started her career at Microsoft, where she spent nearly a decade as a machine learning scientist shaping AI strategies for enterprise products, Kulkarni understands the intricacies of the visa system intimately, having once held an H-1B visa herself. This firsthand experience has fueled her passion for creating a solution that simplifies the immigration process. As Casium continues to grow and innovate, it represents a significant step towards digitizing a critical aspect of the employment landscape, making the journey through the visa maze less daunting for international talent seeking opportunities in the United States.