I'm building a tool that turns messy handwritten notes into clean PDF + LaTeX.

follow along if you’re curious how this evolves and let me know what do you think in comments

working on improving ui + adding diagram support next. still early but here’s what it can already do 👇

#Saas #buildinpublic

Yes, I am experimenting with different OCR models and also Multimodal LLMs. For best accuracy, possibly a hybrid strategy combining a small LLM and OCR model would get quite close. But you're right, it's challenging to get it accurate to the dot. Thanks for your comment 🙂

I'm building a tool that turns messy handwritten notes into clean PDF + LaTeX.

follow along if you’re curious how this evolves and let me know what do you think in comments

working on improving ui + adding diagram support next. still early but here’s what it can already do 👇

#Saas #buildinpublic