Resume Compare using Gemini GenAI API

💡 Inspiration

Resume screening is often time-consuming, subjective, and prone to human error — especially when comparing multiple candidates. We wanted to build a smart assistant that can fairly evaluate two resumes and give instant, AI-powered feedback on who's a better fit for a software engineering position. This tool is inspired by the need for smarter recruitment tools that go beyond keyword matching.

⚙️ What it does

Resume Comparer allows users to upload two PDF resumes and get an AI-generated comparison. Using Google’s Gemini GenAI, it:

Extracts the content from both resumes
Analyzes strengths, weaknesses, and unique qualities
Recommends the more suitable candidate for a software engineering role
Provides a valid, concise justification (50 words)

🛠️ How we built it

Backend: Built with FastAPI to handle file uploads, extract text from PDFs using PyPDF2, and communicate with Gemini AI via the google-generativeai SDK.
Frontend: Developed using React + Vite, styled with Tailwind CSS, and integrated with Axios for API communication.
AI Integration: Used Gemini’s gemini-2.0-flash model for fast and cost-effective text generation, with custom prompt engineering for comparing resumes.

🧱 Challenges we ran into

📄 PDF parsing inconsistencies: Not all resumes extracted cleanly due to different formatting.
🔐 API authentication: Managing and securing the Gemini API key between local/dev environments.
💬 Prompt tuning: Crafting the right prompt to get meaningful, structured, and useful comparisons from Gemini.
⚡ CORS issues between the frontend and backend during local development.

🏆 Accomplishments that we're proud of

🎯 Built a fully functional, AI-powered resume comparison tool end-to-end.
🔗 Seamlessly integrated Google’s cutting-edge GenAI into a FastAPI backend.
🧠 Created a reliable prompt that yields insightful comparisons in natural language.
⚡ Delivered fast performance with a clean, mobile-friendly UI.

📚 What we learned

How to work with Gemini GenAI and prompt it effectively.
Real-world handling of file uploads, PDF processing, and async tasks in FastAPI.
Frontend-backend API communication with CORS and form-data handling.
Deploying and testing AI models in production-ready scenarios.