Resume OCR + LLM HR Analyzer

A simple FastAPI application that accepts a PDF resume, extracts text using PDF text extraction and OCR, then sends the text to a local LLM for HR-style structured analysis.

Features

• Accepts PDF resume uploads
• Extracts text from embedded PDF text or scans pages with Tesseract OCR
• Calls a local LLM endpoint at http://localhost:8000/v1/chat/completions
• Returns HR-friendly structured JSON, summary, and hire/no-hire recommendation

Requirements

• Python 3.12
• tesseract installed on your system
• poppler installed for pdf2image

On macOS:

brew install tesseract poppler

Install

python -m pip install -r requirements.txt

Run

uvicorn main:app --reload --host 0.0.0.0 --port 8001

Example request

curl -F "resume_file=@resume.pdf;type=application/pdf" \
  -F "job_description=We are hiring a senior software engineer with Python and machine learning experience." \
  http://localhost:8001/analyze_resume

LLM API configuration

The app uses these defaults:

• LLM_API_URL=http://localhost:8000/v1/chat/completions
• LLM_MODEL=Qwen/Qwen2.5-1.5B-Instruct

If your LLM requires an API key, set LLM_API_KEY.