AI Resume Parsing System

An end-to-end resume parsing stack with FastAPI backend, Streamlit UI, and MongoDB Atlas storage. It extracts name, email, phone, skills, education, and experience from PDF or DOCX resumes using spaCy NER, regexes, rule-based patterns, and TF-IDF skill matching.

Features

• Upload resumes (PDF, DOCX) via FastAPI or Streamlit.
• Text extraction with pdfplumber and python-docx.
• NLP pipeline with spaCy en_core_web_sm, regex contact parsing, rule-based education/experience extraction.
• Skill detection from a curated dictionary plus TF-IDF cosine similarity.
• Structured JSON storage in MongoDB Atlas with indexes for fast search/filter.
• FastAPI endpoints for upload, retrieval, and search; Streamlit dashboard for manual uploads/search.

Quickstart

1. Python env

   python -m venv .venv
   .venv\Scripts\activate   # Windows
   pip install -r requirements.txt
   python -m spacy download en_core_web_sm

2. Environment

   Copy .env.example to .env and set MONGODB_URI, MONGODB_DB, MONGODB_COLLECTION, optionally SKILL_DICTIONARY_PATH.

3. Run FastAPI

   uvicorn app.api.main:app --reload --port 8000

   • Upload: POST /api/resumes (multipart file file).
   • Get by id: GET /api/resumes/{id}.
   • Search: GET /api/resumes/search?skill=python&text=data&name=smith.

4. Run Streamlit UI

   # from project root
   set PYTHONPATH=.            # PowerShell: $env:PYTHONPATH='.'
   streamlit run app/ui/streamlit_app.py

   Upload a resume to preview parsed JSON and optionally persist to MongoDB.

Notes

• Indexes are created automatically on first DB use (email, skills, name, education.degree, experience.company).
• Skill dictionary lives at data/skills.json; extend it to improve matching.
• For production, secure your MongoDB credentials and consider larger spaCy models.