Matthew

Parseflow - Parse documents into AI ready JSON

by
Parseflow turns PDF, DOCX and TXT files into structured, organized JSON. Extracting useful information while preserving metadata so you can see where all data comes from. Built to be affordable, simple and for devs, small startups and everyday users looking for a way to optimize their workflow. By using Parseflow, you can reduce hallucination and token usage to improve AI models in any environment, personal or commercial.

Add a comment

Replies

Best
Matthew
Maker
📌
Hello Everybody, I'm a student dev, graduating high school this summer. I built Parseflow to help pay for university and also to help AI models understand my school documents. This project started with MagicSchool which is an AI service used by my school, which takes class materiel provided by my teacher and creates AI chatbots to help us with studying. However, it would never give me real answers without me showing my work first and common AI assistants like Gemini, ChatGPT and Claude wouldn't understand my school work properly and give me overcomplicated solutions. So I decided build Parseflow, which takes documents and turns them into structured JSON to help improve context. Once I got my first version, I found I could expand this project to help with more documents like invoices, financials and many others. This way, anybody can use Parseflow to improve their AI's understand of their documents, no matter what use or industry they are in. This brings me to today, where I have completed V2 of Parseflow which is now ready to ship to the public and help everybody. It has two main plans, starter at 9$ month for 500 requests, and growth at 15$ for 1000 requests. Plus BYOK for advanced extraction using your own model providers. There are many more features which you can discover at my website. This product was entirely student built by myself, using fully free tools or free trials from the Github student pack so as this project grows, I will scale with it, in the hopes to some day build a truly useful product for devs and non-devs alike. Any feedback would be greatly appreciated, as well as if you have any questions you can reach out to me at hello@parseflow.tech The project docs can be found at: docs.parseflow.tech