{
"name": "Urvish Babariya",
"role": ["Backend Developer", "Gen-AI Developer"],
"education": "B.Tech ICT @ Dhirubhai Ambani University",
"currentlyBuilding": {
"PharmaChain": "3-tier pharma supply-chain platform, Spring AI compliance copilot (RAG + pgvector + Groq)",
"CodeMentorPRO": "Adaptive AI coding mentor with LangChain + Groq, deployed on Hugging Face Spaces",
"LowLatencyAlgorithms": "Low latency algorithms"
},
"programming": {
"codeforces": "Pupil (max 1360)",
"codechef": "3-Star (max 1612)",
"leetcode": "Knight (max 1956)",
"problemsSolved": "1800+"
},
"learning": ["Low-latency systems", "AI Agents (ReAct)", "Scalable backend architectures", "Generative AI"]
}|
Languages Backend Gen-AI & Agents Databases Tools & DevOps |
|
|
A full-stack pharmaceutical supply-chain platform (React/TypeScript + Spring Boot 3) with PostgreSQL triggers enforcing FDA-style compliance. JWT auth with 5-role RBAC, an append-only SHA-256 cryptographic audit ledger, IoT cold-chain telemetry, and a Spring AI compliance copilot combining RAG (pgvector + Ollama) with Groq tool-calling for natural-language inventory queries. Validated with Testcontainers and shipped via GitHub Actions + Docker.
|
An AI code explainer that adapts to skill level (beginner / intermediate / expert) using Groq-hosted LLMs via LangChain for structured, line-by-line reasoning, deployed through an interactive Gradio interface on Hugging Face Spaces for real-time, adjustable-depth explanations.
|
|
A fully playable, terminal-based Tetris game in C++ built around clean OOP classes for game state and falling blocks — real-time keyboard controls, 2D piece rotation, collision detection, line clearing, and live scoring.
|
|


