Computer Science undergraduate building high-performance web applications, intelligent AI systems, and scalable production-ready software.

🔭 Currently Building: High-performance full-stack applications and intelligent AI-powered systems.

🤖 Focused On: Applied AI, Machine Learning, Temporal Intelligence, Graph-Based Systems, and scalable backend architecture.

🌱 Currently Learning: Advanced ML, LLMs, System Design, Cloud Architecture, and Data Structures & Algorithms.

💼 Experience: Former Full Stack Developer Intern at GoGlobalways, where I contributed to the Bharat AI Olympiad (BAIO) platform.

📄 Publication: Co-Author of a research paper accepted in Elsevier SSRN Proceedings (ICDPN 2026).

🏆 Achievement: Finalist in the Meta × Hugging Face × OpenEnv × PyTorch Hackathon 2026, ranking in the Top 2.5% among 71K+ developers.

🎯 Goal: Build intelligent products and scalable systems that solve meaningful real-world problems.

Temporal Relational Fraud Intelligence Gateway

A real-time fraud intelligence system that combines tabular machine learning with temporal and relational transaction signals.

Traditional fraud detection evaluates transactions individually.

VYUH looks beyond a single transaction.

It combines:

• ⚡ Real-time transaction intelligence
• 🧠 Machine Learning fraud prediction
• 🕸️ Temporal relationship analysis
• 🔗 Bipartite transaction graphs
• 📊 Streaming fraud signals
• 🐳 Containerized deployment architecture

📦 Total Transactions Processed     → 590K+
🧠 Joint Feature Model              → 23 Features
📊 PR-AUC Improvement               → +29.6%
🎯 Fraud Recall Improvement         → +51.2%
📁 Chronological Holdout            → 118K Records
⚡ P50 End-to-End Latency           → 7.46ms
⚡ P95 End-to-End Latency           → 8.38ms

👉 ADD YOUR VYUH GITHUB REPOSITORY LINK HERE

An AI-powered space weather forecasting system for probabilistic GEO electron-flux prediction.

• Fine-tuned Salesforce Moirai-1.0-R-Small
• Used LoRA for efficient model adaptation
• Trained on 11 years of space-weather data
• Processed 1.26M+ five-minute observations
• Combined solar-wind and geomagnetic signals
• Engineered physical wave-derived features
• Used Morlet Continuous Wavelet Transform
• Built inference APIs with FastAPI
• Developed a React-based application layer

📊 Training Data                → 1.26M+ Observations
📅 Historical Data              → 11 Years
📉 Forecast MAE                 → 42.1 → 14.5 pfu
🚀 MAE Improvement              → 65.6%
🎯 Threat Classification        → 94.2%
⚡ CPU Inference                → <5 Seconds

👉 ADD YOUR PRAHARI GITHUB REPOSITORY LINK HERE

Production-Grade MERN E-Commerce Platform

A production-ready MERN e-commerce platform built with a focus on security, payments, real-time functionality, and backend performance.

• 🔐 JWT Authentication
• 🛡️ Helmet.js Security Hardening
• 💳 Razorpay Payment Integration
• 🔒 HMAC-SHA256 Payment Verification
• 📦 Real-Time Order Tracking
• ⚡ Socket.IO WebSockets
• 🛒 Complete E-Commerce Workflow
• 🚀 Production Deployment

🔥 Requests Stress Tested       → 50,000
⚡ Throughput                    → 12,502 RPS
🚀 Average Latency               → 5.86ms
🎯 P99 Latency                   → 12ms
✅ Success Rate                  → 98.5%

May 2026 – June 2026

Worked on the Bharat AI Olympiad (BAIO) platform, contributing across multiple production workflows.

• Built functionality across Student, School, and Admin portals
• Implemented JWT-based Role-Based Access Control
• Developed registration workflows
• Worked on payment workflows
• Implemented result-management systems
• Optimized MongoDB query plans and indexes
• Performed local backend load testing

⚡ Authenticated Reads         → 41.4 RPS
✅ Success Rate                → 100%
🚀 Health-Check Throughput     → 11,494 RPS

🥇 Winner — Devcation Delhi Hackathon 2026, IIT Delhi

🥈 2nd Position — SSH 1.0 Hackathon, Satyaansh SoftTech Pvt. Ltd.

🥉 3rd Runner-Up — InnovAItion Hackathon, Intuitive.ai

🏆 Finalist — Meta × Hugging Face × OpenEnv × PyTorch Hackathon 2026

Ranked in the Top 2.5% among 71K+ developers

📄 Co-Author — Research paper accepted in Elsevier SSRN Proceedings (ICDPN 2026)

🏅 Ranked in the Top 20% of reviewed submissions

👨‍💻 Web Development Lead — Google Developers Group On Campus (GDG OC)

🧩 100+ LeetCode Problems Solved

Co-Author

Accepted in:

🏅 Top 20% of reviewed submissions

⚔️ Building intelligent AI-powered systems
🕸️ Exploring graph-based and relational intelligence
🤖 Learning advanced Machine Learning architectures
⚡ Improving backend performance and system design
📚 Mastering Data Structures & Algorithms
🚀 Shipping impactful real-world products