Sold My AI Study App (1/2)
This will be a two-part series. In this first part, I’ll talk about how I built the app and how I marketed it. In the second part, I’ll talk about the selling process and what I did after the sale.
I sold my app. It took 60 days from the first line of code to the handshake.
This is part one of the story: the build and the hustle. Part two will cover the sale and the aftermath.
You might remember I was working on an AI sidebar called ChatWithYT. It let you “chat” with YouTube videos.
At first, it was basically an Eightify clone. I didn't want to be just another wrapper, so I went looking for a lane. That's when I found the "StudyTok" niche. A few players were dominating:
- StudyFetch
- YouLearn.ai
- TurboLearn.ai
They were all doing the same thing—Whisper-based lecture transcription with some RAG on top—but they were crushing it with students. They had quiz generation, flashcards, and even AI voice agents you could talk to.
I looked at my little browser extension and back at these full-fledged platforms. I decided to pivot.
The Identity Crisis
The problem was, I didn't know who I was building for.
Students? Professionals? Researchers?
The market was split. You had the student-focused tools (StudyFetch, YouLearn), and then the serious research tools (NotebookLM, Eightify). I was stuck in the middle with no clear identity.
The Build
I actually used my own tool. Whenever I hit a boring stretch in a video, I'd use it to get the gist without sitting through the fluff. It worked.
Looking back, I was an idiot. I fell into the classic overthinking trap. I could have sold it right then: "Stop watching boring lectures." Done.
Instead, I wasted a month rewriting the whole thing. I ditched the extension for a full Next.js/FastAPI web app. I added every feature I could think of:
- Model switching (Claude, Gemini, DeepSeek—the works)
- A Notion-style editor with AI autocomplete
- File imports for every format under the sun
- Live lecture recording
- Auto-generated study guides and flashcards
The result was a Frankenstein product sitting awkwardly between "simple utility" app and full-fledged platform.
The Hustle
We went all in on TikTok.
I bought six iPhones and subscribed to a couple of VPNs (HMA and Mullvad). My girlfriend and I spent our evenings creating content.
We didn't invent anything. We just stole winning formats from US creators and localized them for a Vietnamese audience.
Why Vietnam? I'll get to that.
The trick is volume. Once you find a hook that works, you beat it to death. You remix it, repost it, and milk it until it stops performing.
It worked. We hit 200 users in under three weeks.
From Building to Selling… in 60 Days?!
Here's the twist: I built this for Vietnam because I already had a buyer.
I met him before I even wrote a line of code. I pitched him the concept, and he was interested.
We met for coffee in Ho Chi Minh City, hopped on a few calls, and agreed on a number.
So, I spent a month frantically building the core features we discussed. Once it was ready, we met again to sign the deal.
(Part 2 coming soon.)