Skaff
An AI-powered candidate intelligence platform that transforms resumes and GitHub profiles into interactive skill insights.
Overview
Skaff is an AI-powered candidate intelligence platform built to help recruiters, HR teams, and students better understand technical skills beyond what's written on a resume. By combining resume parsing, GitHub analysis, AI-powered skill extraction, and interactive 3D knowledge visualization, Skaff transforms scattered technical information into meaningful hiring insights.
Originally developed during HACK KRMU 5.0, the project qualified for Round 2, demonstrating the viability of using AI and data visualization to improve technical recruitment workflows.
Problem
Traditional hiring often depends on manually reviewing resumes, GitHub profiles, and portfolios separately. This process is time-consuming, inconsistent, and makes it difficult to understand a candidate's actual technical strengths, technology relationships, and overall suitability for a role.
Recruiters frequently compare candidates based on keywords instead of real technical depth, while students struggle to visualize their own strengths and identify missing skills.
Skaff was built to simplify both experiences.
Solution
Skaff analyzes resumes and GitHub repositories to automatically identify programming languages, frameworks, libraries, and technical skills using AI-powered natural language processing. The extracted information is transformed into an interactive 3D knowledge graph where technologies are organized by relationships and dominance.
Recruiters can compare two candidates side-by-side, view AI-generated hiring insights, identify missing skills, evaluate technology dominance, and receive role recommendations through a unified interface.
Students can better understand their own technical profile while recruiters gain a faster and more structured hiring workflow.
Architecture
Skaff uses React and React Three Fiber to render an interactive 3D visualization layer, while FastAPI provides backend services for GitHub analysis, resume processing, AI inference, and graph generation.
Gemini API performs natural language processing to extract technical skills and generate hiring insights. Supabase manages authentication, PostgreSQL storage, and role-based access control across Student, Recruiter, HR, and Administrator roles.
The platform follows a modular architecture with dedicated services responsible for GitHub analysis, resume parsing, AI recommendations, role matching, and knowledge graph generation.
Engineering Decisions
AI-assisted skill extraction instead of keyword matching.
Rather than relying solely on keyword detection, Skaff combines structured GitHub data with AI-powered resume analysis to identify technologies, relationships, and candidate strengths more accurately.
Knowledge Graph visualization over traditional dashboards.
Instead of presenting skills as static lists or charts, the platform visualizes technical relationships through an interactive 3D graph, making complex information easier to explore and compare.
Role-based platform architecture.
Students, Recruiters, HR teams, and Administrators each receive interfaces designed specifically for their workflows while sharing a common backend infrastructure and authentication system.
Modular AI services.
Resume analysis, GitHub parsing, graph generation, and recommendation engines operate independently, making the platform easier to extend and maintain as new features are introduced.
Challenges
The most challenging part of Skaff was combining multiple independent systems into a cohesive workflow. Resume parsing, GitHub repository analysis, AI-powered skill extraction, authentication, and 3D visualization all needed to operate together while maintaining acceptable performance.
Designing an intuitive knowledge graph also required balancing visual clarity with technical depth. Technologies needed to be grouped logically, interactive enough for exploration, and detailed enough to support recruiter decision-making without overwhelming users.
Additionally, working within Gemini API rate limits during development required optimizing API usage and restructuring requests to maximize available quota.
Impact
- Qualified for Round 2 of HACK KRMU 5.0 Hackathon.
- Built an AI-powered platform combining resume parsing, GitHub analysis, and candidate comparison.
- Implemented interactive 3D knowledge graph visualization using React Three Fiber.
- Developed role-based dashboards for Students, Recruiters, HR teams, and Administrators.
- Integrated Gemini AI for technical skill extraction, candidate summaries, and hiring recommendations.
- Built secure authentication and cloud-based data management using Supabase.
Product Vision
Skaff is designed to move technical hiring beyond keyword matching and resume screening.
The long-term vision is to build an intelligent hiring platform capable of evaluating technical candidates through structured skill relationships, AI-assisted reasoning, coding profiles, portfolio analysis, certifications, and future assessment integrations.
Rather than replacing recruiters, Skaff aims to reduce manual effort and provide deeper technical insights that support faster, more informed hiring decisions.
What I Would Improve
Future development will focus on expanding GitHub repository analysis, integrating additional coding platforms such as LeetCode and Codeforces, supporting multiple AI providers, improving recruiter analytics, and enhancing recommendation accuracy through richer technical signals.
I also plan to introduce collaborative hiring workflows, organization-level dashboards, advanced graph analytics, and scalable graph database integration using Neo4j to support larger candidate datasets.