AI Resume Evaluator & Ranker

Workflow Diagram
Input Screen
Results Table
Download Feature

About the project

Inspiration

Recruiters often spend hours manually screening resumes, which is both time-consuming and subjective. We wanted to build a tool that makes this process faster, fairer, and more data-driven.

What we learned

While working on this project, we deepened our understanding of Natural Language Processing (NLP), especially techniques like TF–IDF vectorization and cosine similarity. We also learned how to structure a project for usability—designing a system that is simple for recruiters but powerful under the hood.

How we built it

We built a pipeline where resumes and job descriptions are first converted into numerical vectors using TF–IDF.
We then calculated cosine similarity to score how well each resume matches the given job description.
Finally, we created a clean, interactive interface where recruiters can upload resumes, see ranked results instantly, and download the data for further use.

Challenges we faced

Handling resumes in different formats (PDF, DOCX) and ensuring clean text extraction.
Designing the similarity scoring so that it balances keywords with overall context.
Making the app efficient enough to process multiple resumes without lag.

Through these challenges, we learned how to combine AI models with practical software engineering to solve a real-world problem.