DocuMint - AI Document Extractor

InspirInspiration:

Managing invoices, contracts, resumes, and other documents manually takes time and makes it difficult to find important information. I wanted to build a smarter solution that could automatically extract useful information from documents and organize it in one place. This idea inspired me to create DocuMint, an AI-powered document extraction and management platform.

What it does:

DocuMint allows users to upload PDF documents and extract structured information using AI. It organizes the extracted data in a searchable dashboard and supports exporting results to CSV and JSON formats. The project uses the Nebius Token Factory API to connect with an AI model for document understanding.

How I built it:

I built the backend using Python and FastAPI, with SQLAlchemy for database operations. PyMuPDF helps extract text from PDFs, while a rule-based extraction method provides a fallback when AI extraction is unavailable. I integrated the Nebius Token Factory API using an OpenAI-compatible client. The frontend uses React and Vite to provide a simple interface for uploading, searching, viewing, and managing documents.

Challenges I faced:

One of the main challenges was converting unstructured document text into consistent, structured data. I also had to handle different PDF formats, API configuration, missing fields, and errors during extraction. I addressed these challenges by combining text extraction, AI-powered processing, validation, and fallback logic.

What I learned

Through this project, I learned how to integrate AI APIs into a full-stack application, build REST APIs with FastAPI, manage data using a database, and connect a React frontend with a Python backend. I also learned about API key security, error handling, and the importance of checking AI-generated results.

Future improvements

Future improvements include adding user authentication, secure cloud storage, OCR for scanned documents, support for more file formats, and improved extraction accuracy for different types of invoices and constatation.

Built With

• ai
• css
• fastapi
• git
• html
• javascript
• llm
• nebius-token-factory
• pymupdf
• python
• react
• rest-api
• sqlalchemy
• sqlite
• vite