Paul Jung

Paul Jung grew up in a low income household as the child of first generation immigrants, went through the UK state school system, and could not afford the private tutoring that wealthier classmates used to climb the ladder. That frustration followed him into medicine, where he met cofounder Kavi Samra at UCL, and both saw the same pattern again and again in the exam treadmill that controls access to universities and professional careers.

Instead of staying in secure NHS jobs, they left medicine and pointed their training toward a single question: could an AI tutor reproduce the best parts of a top private tutor, but at a price point and availability that works for every GCSE, A level, IGCSE and IB student, not just the top decile. Medly AI was the answer, built as an AI first exam prep platform aimed squarely at high stakes tests that parents already overspend on, with Paul as CEO coordinating product, pedagogy and regulatory proof.

Medly launched publicly in February 2025 and focused ruthlessly on UK exam preparation markets that already have clear demand and tight feedback loops, starting with GCSE, IGCSE, A level and International Baccalaureate exams. Within roughly nine months they reported more than 400,000 students using the AI tutor in the UK alone and a 600 percent increase in the user base since launch, hard numbers that turned vague AI edtech narratives into something investors could underwrite.

Instead of selling only a story about access, Jung stacked proof that Medly changes outcomes and can work at national scale. The company secured a highly competitive Department for Education AI Tutoring Tools Pioneers Programme grant, ran a randomized controlled trial across eight UK schools to show measurable grade improvements, and then used that evidence and retention data to raise an $8 million seed round led by Felix Capital with Ada Ventures and Eka Ventures returning, which gave them both capital and validation from specialist early stage funds.

1. Narrow the problem to a money backed pain, then instrument it. Jung chose exam preparation where parents already spend thousands on tutoring, constrained the product to specific exam boards, and built Medly to slot directly into revision workflows, which made it easy to test whether students were actually learning more and to track retention and engagement instead of just signups.

2. Use public sector credibility as a growth wedge. By winning the Department for Education AI Pioneers grant and running a randomized controlled trial across eight schools, Medly gained something few consumer edtech apps have: government backed proof that the tool does not just keep kids busy but moves grades, which made it easier to convince both schools and investors that this was infrastructure, not a gimmick.

3. Turn retention into the fundraising story. When Jung announced the $8 million seed, the headline metric was that 75 percent of students were still active three months after sign up, and that more than 400,000 students had already used the app, so the pitch to funds like Felix Capital and Ada Ventures was not that AI tutors are a hot theme, it was that Medly had already cracked the hardest part of consumer learning products, getting students to come back long enough for outcomes to compound.