Role Radar
Discover professionals with 10+ years of experience in any domain using NLP and real-time web scraping.
Duration
2 weeks
Team
4 members
Accuracy
100%
Technologies Used
Project Overview
Role Radar is a Flask-based web application that leverages NLP and real-time data scraping to identify domain experts with over a decade of experience. By taking minimal user input (domain and optional keywords), it intelligently fetches and filters profiles from GitHub and search engines, ranking them using a custom-built confidence scoring algorithm.
Methodology
Challenges
Scraping reliable data in real time without being rate-limited
Designing a confidence score that balances multiple relevance factors
Integrating multiple APIs seamlessly within response time goals
Solutions
Used Serper.dev for Google-like search results and GitHub API for detailed metadata
Developed a Gemini-powered NLP algorithm to evaluate context and experience
Streamlined backend requests to ensure <20s total response time
Results & Impact
Returned top experts in under 20 seconds per search
Confidence score accuracy validated with test prompts
Effortless UX with no dummy data; 100% real profiles
Interested in This Project?
Want to learn more about the implementation details or discuss similar projects? I'd love to hear from you!