job-application-ai-assistant

Leveraged MCP and OpenAPI to implement a multi-agent job application AI assistant

Overview

This project is an agentic workflow orchestration system built with the Model Context Protocol (MCP) on FastAPI. It automates job application processing using ChatGPT for job description summarization and OpenAI embeddings for semantic similarity scoring. It also manages email notifications with the Resend API and stores data and error logs in SQLite.

Features

• Resume Extraction Extracts text from PDF and DOCX files using PyMuPDF and python-docx
• Job Description Summarization Generates concise summaries using ChatGPT (gpt-3.5-turbo)
• Semantic Scoring Computes cosine similarity using OpenAI Embeddings (text-embedding-ada-002) and numpy
• Email Automation Sends interview invitations via the Resend API for candidates above threshold
• Duplicate Detection Prevents redundant processing by checking existing applications by email, resume, and job description
• Resume Validation Validates documents as resumes using ChatGPT-based analysis
• Error Logging Records errors in a SQLite database for auditing
• Custom MCP Server Exposes tools as reusable endpoints under /mcp
• Responsive Web UI Modern Bootstrap interface for easy form submission

Architecture

1. FastAPI serves as the HTTP server
2. MCP extends FastAPI with agentic tools under /mcp
3. Database layer uses SQLAlchemy with SQLite for storage
4. File handling saves uploaded resumes in an uploads directory

Technical Stack

Component	Technology	Purpose
Framework	FastAPI	API server
MCP	fastapi_mcp (stub)	Agent orchestration
Database	SQLAlchemy, SQLite	Persistent storage
NLP	openai	ChatGPT summarization and embeddings
Email	resend	Notification delivery
File Processing	PyMuPDF, python-docx	Resume text extraction
Utilities	numpy	Numerical operations for similarity scoring

Dependencies

• fastapi
• uvicorn
• python-dotenv
• openai >= 1.0
• resend
• sqlalchemy
• PyMuPDF
• python-docx
• python-multipart
• numpy
• fastapi_mcp (included)

Usage

1. Obtain API keys for the external services:
   • Sign up at OpenAI and create an API key for OPENAI_API_KEY.
   • Create an account at Resend to generate a RESEND_API_KEY.
2. Copy .env.example to .env and replace the placeholder values with your own keys. The .env file is ignored by Git so your keys remain private.
3. Create and activate a virtual environment (recommended):

   python -m venv venv
   source venv/bin/activate

4. Install the required Python packages using your environment's Python:

   python -m pip install -r requirement.txt

   You can also run ./setup.sh, which installs the same dependencies. If you see ModuleNotFoundError errors when running the app, rerun the above command inside the same virtual environment to ensure all dependencies are installed.
5. Ensure the uploads directory is writable.
6. Run python main.py to start the API server (listens on http://127.0.0.1:8000).
7. Open http://localhost:8000 in your browser to access the Bootstrap-based web UI and API.

API Endpoints

POST /applications

Processes a job application by analyzing the resume and job description

• Parameters file (PDF or DOCX) job_description_text
• Response JSON with email, score, email_status, message

MCP Tools

Tools are available under /mcp for individual testing:

• extract_resume_text
• summarize_job_description
• score_similarity
• extract_email_address
• send_email_notification
• invite_for_interview
• find_existing_application
• validate_resume_document

Database Schema

• applications id, email, resume_text, job_description, score, email_status, created_at
• error_logs id, error_message, created_at

Notes

• Use OPENAI_API_KEY for OpenAI access
• Customize the OPENAI_MODEL environment variable to switch chat model. The sample .env.example uses gpt-4.0 as the default model.
• The fastapi_mcp module is bundled with the repository as a lightweight stub, so there is no external fastapi-mcp package to install.
• This project requires the openai Python library version 1.0 or newer. If your environment has an older version installed, upgrade with pip install -U openai.