🎉 Big News — I just launched my own AI product!
I finally built something I’ve always wanted for myself —
an AI tool designed to help people learn from YouTube more effectively and build their own knowledge base.
It’s called LearnTube AI.
It started as a simple tool I made for myself to study more efficiently on YouTube — to take better notes, organize what I learn, and turn videos into structured knowledge.
When I mentioned it during my talk at Imperial College last weekend, many people showed strong interest.
So I interviewed 10+ users, collected feedback, iterated quickly… and now the product is officially live! 🚀
Here’s what LearnTube AI can do:
🌏 Auto subtitles & translation – extract video transcripts and bookmark or save specific lines
🕒 Adjustable timeline – fine-tune subtitle timing and jump to key moments
🧠 AI Summary – generate concise, multi-language summaries of each video
📚 Smart insights – detect and explain key terms, apps, and concepts mentioned in the video
🔈 Audio mode – listen to AI summaries (the voice is still a bit robotic, but improving! 😅)
🖍️ Highlight & Notes – mark key ideas, write reflections, and build your own knowledge library
This project was completely vibe-coded by me from scratch — from design to development to launch.
I ran into plenty of bugs (and some tears 😭), but when I saw it actually working for the first time, the sense of accomplishment was incredible.
I also tested nearly every major “vibe coding” tool out there during this process and even wrote a comparison article — a fun little bonus outcome.
LearnTube AI is now live on both desktop and mobile.
If you love learning from YouTube but don’t have time to take detailed notes, this might be exactly what you need.
💚 Try it here: https://learntubeai.ai
I’d love to hear your thoughts, feedback, or ideas.
Thank you all for the support and encouragement along the way ❤️
If you want to arrange a call for any questions or conversations, send me an email: jyang1002@gmail.com or find me on LinkedIn.
HOW IT STARTED
LearnTube AI began as a small side project — a tool I originally built for myself.
I’ve always learned new topics from YouTube, but I often found it hard to review what I’d watched or remember key points.
So I created a simple prototype that could extract a video’s transcript and summarize it automatically.
When I shared it with a few friends, they were immediately interested and asked if they could try it.
Their feedback helped me realize that this wasn’t just my problem — it’s a universal challenge for anyone learning from videos.
I invited about a dozen people to test the early version, collected their feedback, and started iterating.
Through this process, the product evolved from a personal learning assistant into a public-facing tool that helps anyone learn faster from video content.
