Skip to main content

← All case studies

In development

PdfAnalysis: document extraction platform

A document platform that turns PDFs into structured JSON, Excel, and CSV using templates, tables, OCR where needed, and a rule-based path that does not depend on an LLM.

Discuss a similar project
Status
In development
Category
Document systems
Stack
Astro · Vue · FastAPI
Client
PdfAnalysis
Document extraction platform

Business problem

Teams were extracting fields from PDFs by hand. They needed repeatable templates, bulk jobs, and an API, not a chat window around a file.

What needed to happen

CodeForge product. The site URL is stored as intended only. Functionality is not live, so this is shown as in development, not as a launched or production site. LLM steps exist as optional helpers, not the product definition.

What I built

Uploads are processed through template detection, table and text extraction, and optional OCR. A rule-based engine covers common fields without an LLM. Bulk jobs, API keys, webhooks, and credits sit around that core.

How it works

A document extraction product with templates, structured output, and APIs. Shown as in development. No View Live Site while functionality is not live.

Technical detail

  • Astro and Vue frontend; FastAPI, Celery, Redis, and Supabase on the backend
  • PyMuPDF, Camelot, and Tesseract for PDF, tables, and OCR
  • Rule-based extraction for the free path; optional LLM mapping only where configured
  • Stripe billing, API keys, and webhooks for programmatic use

What shipped

  • PDF upload to structured JSON, Excel, and CSV
  • Templates and bulk jobs
  • REST API and webhooks
  • Rule-based extraction that does not require an LLM

Technology

  • Astro
  • Vue
  • FastAPI
  • Celery
  • Redis
  • Supabase
  • Stripe
  • PyMuPDF

Discuss a similar project

Short enquiry first. A full brief is available if you already have the detail.

Discuss Your Project