SkillScan

Inspiration

As we are all applying to the intership, the struggle of our resumes going unseen is a pain. This helps not only us but also recruiters by improving productivity.

What it does

SkillScan is an automated resume evaluation tool designed to help employers manage the thousands of resumes they receive for a single position.

It replaces the repetitive manual work of evaluating every candidate by providing a ranked list of top applicants based on specific employer needs.

Users can specify parameters like leadership, experience, education, and hard skills to filter for exactly what they want to see.

How we built it

The system allows users to input a folder of resumes and define their desired candidate profile.

We built a pipeline that transforms .pdf files into Base64.

This data is sent via API to Google Gemini, which uses specific prompts and a grading metric to evaluate the candidates.

The final output is a ranked page showing only the top number of candidates requested by the employer.

Challenges we ran into

Technical Limitations: We had to navigate API request constraints and processing speeds.

User Input Complexity: We found that as the amount of required user input increases, the likelihood of the system running into processing issues also rises.

Accomplishments that we're proud of

Successfully automating a traditionally manual and "repetitive" task.

Integrating Google Gemini to handle complex resume grading rather than just simple keyword matching.

Creating a tool that can rank candidates accurately based on diverse metrics like "hard skills" and "leadership."

What we learned

We learned how to manage PDF-to-data transformations for AI processing.

We gained experience in prompt engineering to ensure the AI evaluates resumes consistently against a specific grading metric.

We recognized the balance between offering user customization and maintaining system stability.

What's next for SkillScan

Expanded Evaluation: We plan to include new categories like extracurriculars, soft skills, and a fully customizable "Additional" section.

Direct Links: We want to provide direct links to the original resume PDFs under each candidate's name in the ranking.

Performance: Improving API request handling and speed to make the tool faster for high-volume users

Built With

• css
• geminiapi
• html
• next.js
• tailwind
• typescript