Alice Turned Exam Prep Into Something Students Actually Want to Use

The Macro: Ed-Tech Has a Credibility Problem

The ed-tech space has produced a lot of products that look great in pitch decks and fall apart in actual classrooms. Quizlet has been around for two decades and still feels like digital flashcards. Chegg built a business on homework answers that professors openly despise. Coursera and Udemy are content libraries, not learning tools. The gap between “technology for education” and “technology that makes students learn better” is wide and mostly unfilled.

Duolingo is the exception that proves the rule. It works because it understood something fundamental: learning is a retention problem, not a content problem. The content is almost secondary. What matters is the system that gets you to come back every day, practice the right things at the right time, and feel like you are making progress. Gamification gets mocked by serious educators, but Duolingo’s results speak for themselves.

The university exam prep market is different from language learning, though. Students are not trying to build a long-term skill. They are trying to pass a specific exam in three weeks using specific course materials that their specific professor assigned. The inputs are messy: lecture slides, textbook chapters, handwritten notes, random PDFs. The output needs to be targeted: practice questions that mirror what will actually be on the test.

AI changes the math here because it can do something no static tool could: take arbitrary course materials and generate structured learning content from them. Flashcards, practice tests, mind maps, all tailored to what this particular student needs to study for this particular exam. That is a product that could not have existed two years ago.

The Micro: A McKinsey Partner Who Actually Taught Classes Built This

Alice is an AI-powered study platform that transforms uploaded course materials into personalized exam prep. You upload your PDFs, lecture slides, or notes, and the system generates mind maps, flashcards, multiple-choice questions, and exam simulations based on that content. The pitch is “Duolingo for exams,” which is a comparison that actually holds up when you look at what the product does.

The feature set is more comprehensive than most ed-tech tools I have seen. Mind mapping automatically organizes scattered study materials into connected course overviews. The practice tools generate questions at varying difficulty levels. There is a motivational system called “Carrot Journey” that tracks progress, which sounds gimmicky but is probably necessary for a product targeting undergrads who would rather do anything than study. AI chat answers course-related questions, and study groups enable collaborative learning with shared notes.

The traction numbers are real. Over 100,000 university students use the platform, including students at MIT, Stanford, CBS, and Berkeley. They are at $20K MRR with 10% week-over-week growth. Those are strong numbers for an ed-tech product that is less than a year old.

Kim Rants is CEO and co-founder. His background is what makes this company interesting. He was a McKinsey Associate Partner specializing in AI, the fastest-promoted engagement manager in Europe. He was Head of Global Business Development at LEGO. And he has eleven-plus years of university teaching experience. That last detail matters more than the consulting pedigree. Most ed-tech founders have never stood in front of a classroom. Kim has done it for over a decade. He has an MSc in Computer Science from Imperial College London and an MSc in Finance from LSE.

Patrick Gadd is CTO and co-founder. He was a Senior Data Scientist at LiveIntent, a US ad-tech company, where he developed a patented ML solution for audience targeting. He has an MSc in Computer Science from Oxford. They have grown the team to five people, based in Copenhagen, and are part of YC’s Winter 2025 batch.

The competitive space includes Quizlet, Anki, Studocu, and Brainscape for flashcards and study tools. Photomath and Socratic handle specific subject areas. But none of these take arbitrary course materials and generate a full exam prep experience from them. That is the differentiator.

The Verdict

I think Alice has something real. The combination of genuine teaching experience in the founding team, strong early traction, and a product that solves a specific problem (turning messy course materials into structured exam prep) is compelling. The 10% week-over-week growth suggests students are telling other students about it, which is the only distribution channel that works in education.

The risks are seasonal and structural. University exam prep is cyclical. Usage will spike before midterms and finals and drop during breaks. Building a business on seasonal demand is possible but it requires careful financial planning. The structural risk is that universities might build or buy this capability themselves, integrating AI study tools directly into their LMS platforms like Canvas or Blackboard.

Thirty days, I want to see retention through a non-exam period. Do students keep using Alice between test cycles, or does it go dormant? Sixty days, show me the expansion into new universities. Is adoption spreading virally across campuses, or does each new school require a push? Ninety days, the question is pricing power. Students are famously cheap. Can Alice charge enough per student to build a real business, or does it need to flip to a B2B model and sell to universities directly? The $20K MRR at 10% WoW growth puts them on a good trajectory, but the path from “students love it” to “sustainable business” has tripped up a lot of ed-tech companies before.