Kushal
Rongali

Engineered a full-stack real-time collaborative coding platform using the MERN stack and WebSockets, enabling multiple users to write and execute code simultaneously in shared sessions.

• Architected isolated execution environments per user session — secure, independent code runs with zero cross-session data leakage
• Room-based collaboration with host controls, access management & live participant tracking for structured pair-programming workflows
• Optimised real-time code synchronisation achieving low-latency bidirectional communication, keeping editor state consistent across all connected clients under concurrent load

Production-grade, multi-role SaaS platform (Student / University / Recruiter) built with a monorepo MERN stack — React + TypeScript frontend served by a single Express backend deployed on Render.

• Integrated OpenAI/Groq API for AI mock interviews with rubric-based scoring across 6 dimensions (Technical Depth, Problem-Solving, Communication, etc.) and automated resume parsing from PDFs
• Built GitHub code analysis engine fetching repo metadata, commit history & language distribution to generate an AI Engineering Score; integrated LeetCode stats scraper for holistic Placement Readiness scoring
• Dual-token JWT auth (15-min access + 7-day refresh), role-based access control across 3 roles, rate-limited REST API (100 req/15 min), and Helmet security headers for production hardening
• Designed 18 Mongoose schemas covering candidate pipelines, interview scheduling, skill passports, AI roadmaps, and university cohort analytics with batch CSV onboarding for institutional users