Opennote Built an AI Tutor That Lives Inside Your Notes

The Macro: Notes Apps and AI Tutors Exist, but Nobody Combined Them Well

The note-taking market is crowded and mature. Notion has over 30 million users. Obsidian has a fanatical following among knowledge workers. Roam Research pioneered bidirectional linking. Google Docs and OneNote are installed on basically every computer on earth. Students have more options for writing things down than at any point in history.

Separately, the AI tutoring space exploded after ChatGPT launched. Khanmigo from Khan Academy. Duolingo Max. Photomath. Quizlet’s Q-Chat. Dozens of startups building AI study assistants that answer questions, generate flashcards, and explain concepts.

The problem is that these two categories exist in parallel and almost never intersect. You take notes in one app. You ask questions in another. The AI tutor has no idea what you wrote down in class. Your notes app has no idea what you are struggling with. The result is that students bounce between six different tools and none of them have the full picture.

This gap is real and I have watched students work through it. They copy text from their notes into ChatGPT. They paste the answer back into their notes. They screenshot a diagram and upload it to an AI app. They manually create flashcards from material they already wrote down once. Every step is friction. Every transition loses context.

The opportunity is obvious: put the AI inside the notes. Not as a sidebar feature or a plugin. As a first-class part of the notebook experience.

The Micro: Three Founders, 55,000 Students, and an AI Named Galileo

Opennote was founded by Rishi Srihari, Abhi Arya, and Vedant Vyas. They came through Y Combinator’s Summer 2025 batch and are based in San Francisco. Three-person team, which is standard for this stage.

The product is a notebook with an AI thinking partner called Galileo built directly into the interface. You take notes. Galileo reads them. When you have a question, Galileo answers it with full context of everything you have written. When you need to study, Galileo generates flashcards and quizzes from your actual notes, not generic material. When you record a lecture or paste a YouTube link, Galileo converts it into structured notes and summaries.

I like that Galileo uses Socratic questioning rather than just dumping answers. There is a meaningful pedagogical difference between an AI that tells you the answer and one that asks you the right questions until you figure it out yourself. Whether most students actually prefer the Socratic approach over “just tell me” is an open question. My guess is that the students who stick around long enough to appreciate it become the most loyal users.

The feature set is broader than I expected for a team this size. Rich text editor. AI chat. Audio recorder with transcription. Tutorial generation including videos and diagrams. Practice tools with inline flashcards and quizzes. That is a lot of surface area. The question is whether they can maintain quality across all of it.

The traction number is impressive: 55,000 students. For a product from a three-person startup, that is real adoption. Student testimonials on the site suggest genuine enthusiasm. One engineering student said Galileo explains things better than his textbook. A graduate student said he barely uses other document tools anymore.

Competitors in this specific intersection are thin. Notion AI exists but it is a general-purpose writing assistant, not a study tool. Reflect and Mem have AI features but target knowledge workers, not students. Quizlet is a flashcard app that added AI, not a notebook that added learning. Opennote is coming at this from the notebook side, which I think is the right entry point because notes are where learning starts.

The Verdict

I think Opennote is sitting on something with real potential. The 55,000 user number tells me students want this. The product design tells me the team understands how learning actually works, not just how note-taking works.

The risk is that this is a consumer education product with a freemium model. Monetizing students is notoriously difficult. They are price-sensitive, they churn when semesters end, and their willingness to pay correlates more with exam stress than with product quality. The best student products either convert to professional tools after graduation or find institutional buyers like universities.

The other risk is that Notion or Obsidian ships a study mode. Both companies have the distribution and the AI infrastructure to build this feature in a quarter. Opennote needs to get deep enough into the student workflow before that happens.

In 30 days I want to see retention by cohort. Are students who start using Opennote in February still active in April? In 60 days the question is whether collaborative features are driving organic growth. Students telling other students is the only acquisition channel that works at scale in education. In 90 days I want to know if universities are reaching out. Institutional contracts are where education startups go from interesting to defensible. The notebook that thinks with you is a great pitch. Now it needs to prove it can also grow with you.