America’s visa system is notoriously complex, and Priyanka Kulkarni knows it better than most. The 34-year-old machine learning scientist spent nine years juggling the visa process herself. Now, she’s using artificial intelligence to help others handle work-based immigration smoothly and hassle-free.
Who is Priyanka Kulkarni?
Kulkarni grew up in India and joined Microsoft right after college. She spent ten years as a machine learning scientist on an H-1B visa. But the visa process was stressful. “Honestly, it was exhausting, confusing, and at times very career-limiting,” she told Business Insider. When the firm offered her a spot in 2024, she applied for an EB-1 visa. It took three months with a law firm to complete the paperwork. On her first day at the new company, when asked what she wanted to build, she said, immigration tech. “Everything I’ve done has led to this point,” she said.
How Priyanka Kulkarni is using AI to transform immigration
According to a report by Business Insider, Casium, founded by a former Microsoft data scientist, aims to make life easier for immigrants working in the USA. The platform gives companies a way to handle visa applications from start to finish. It successfully replaces Excel sheets and, in many cases, expensive outside law firms. The product is designed with the constantly changing immigration scene in mind, including policies that are often difficult to track and subject to frequent changes. For instance, the Trump administration’s surprise rule required companies to pay $100,000 for every new H-1B visa. While some firms accepted it, others protested, and lawsuits followed. Even with the H-1B, there are exceptions, such as the rule only applying to new filers and not existing ones, and many of these regulations can confuse or be misunderstood by people.
Casium saw the problem and now they are using technology to make the process faster and easy. The company says it has helped hundreds of candidates with assessments, compliance checks, and filings and in return got “very high approval rates.” In some cases, hires have started work less than a month after signing up.
Casium was founded in 2024 and has recently raised $5 million in seed funding from Maverick Ventures, Ai2 Incubator, GTMfund, Success Venture Partners, and angel investor Jake Heller, according to Business Insider. The startup did not reveal its valuation.
How Casium works
In an interview with the Business Insider, Kulkarni revealed that the process involves a candidate filling out the form first. Next, a set of AI “agents” looks at public data, patents, and research papers to understand the candidate’s profile. Within the next few minutes, Casium recommends the best visa like H-1B, O-1, or EB-1A.
The analysis is then sent to licensed lawyers and paralegals working with Casium. With one click, a draft letter reveals whether the candidate’s eligibility is ready. According to Kulkarni, this tech reduces paperwork time from three to six months with a traditional law firm to less than ten business days.
