███████╗██████╗ █████╗ ███╗ ██╗██╗ ██╗██╗ ██╗███╗ ██╗ ██████╗ █████╗ ██████╗ ██╗ ██╗
██╔════╝██╔══██╗██╔══██╗████╗ ██║██║ ██╔╝██║ ██║████╗ ██║ ██╔══██╗██╔══██╗██╔══██╗██║ ██║
█████╗ ██████╔╝███████║██╔██╗ ██║█████╔╝ ██║ ██║██╔██╗ ██║ ██████╔╝███████║██████╔╝██║ ██║
██╔══╝ ██╔══██╗██╔══██║██║╚██╗██║██╔═██╗ ██║ ██║██║╚██╗██║ ██╔══██╗██╔══██║██╔══██╗██║ ██║
██║ ██║ ██║██║ ██║██║ ╚████║██║ ██╗███████╗██║██║ ╚████║ ██████╔╝██║ ██║██████╔╝╚██████╔╝
╚═╝ ╚═╝ ╚═╝╚═╝ ╚═╝╚═╝ ╚═══╝╚═╝ ╚═╝╚══════╝╚═╝╚═╝ ╚═══╝ ╚═════╝ ╚═╝ ╚═╝╚═════╝ ╚═════╝
class FranklinBabu:
name = "Franklin Babu"
handle = "@Franklin1312"
status = "Student who likes to build things"
location = "India"
stack = {
"AI/ML" : ["XGBoost", "LightGBM", "scikit-learn", "FastAPI", "Python"],
"Blockchain" : ["Solidity", "Ethers.js", "Hardhat", "Sepolia ETH", "Web3.js"],
"Frontend" : ["React", "JavaScript", "HTML5", "CSS3", "PWA"],
"Backend" : ["Node.js", "Express", "MongoDB", "REST APIs"],
"Tools" : ["Git", "VS Code", "Hardhat", "Vercel", "Render"],
}
currently_building = "AI-powered parametric insurance for gig workers"
started_with = "A knockout tournament bracket in Python — in 12th grade"
fun_fact = "My ML model hit R2 = 0.9874 on the first real dataset I trained"Protecting India's 12 million gig delivery workers from income loss during disasters
When a flood, AQI spike, or curfew hits, KAVACH's AI predicts exactly how much a delivery worker would have earned — and pays them automatically within minutes via UPI. No forms. No waiting.
What makes it extraordinary:
- 🧠 Stacked ML ensemble (XGBoost + LightGBM + Ridge) — R² 0.9874, MAE ₹16.79, trained on 60,000 samples
- 🔗 Blockchain audit trail — every payout hashed & logged immutably on Sepolia ETH (Etherscan-verifiable)
- 🚦 7-layer fraud engine + Physical Presence Confidence Score (PPCS) — detects GPS spoofing rings
- 📡 Multi-source parametric triggers — rain, AQI, flood, curfew, platform outage across 16 Indian cities
- 💸 Zero-touch payouts via Razorpay UPI within minutes of disruption detection
- 🌍 Multilingual PWA — English, Hindi, Tamil with Claude AI chatbot built-in
Stack: React · Node.js · FastAPI · Python · MongoDB · Solidity · Ethers.js · Razorpay · Claude API
Farm to consumer transparency — eliminating middlemen, empowering farmers
A decentralized platform where farmers create product batches on-chain, transporters update states, retailers purchase with automated payment splits — and consumers scan a QR code to see the full journey of their food.
Key features:
- ⛓️ Smart contract on Sepolia ETH with role-based access (Farmer / Transporter / Middleman / Retailer / Consumer)
- 💰 Automated payment distribution on purchase — no middleman can intercept farmer pay
- 📱 QR code traceability — consumers see every hand the product passed through
- 🔒 OpenZeppelin AccessControl + ReentrancyGuard — production-grade security
Contract: 0x10DA8a6Ff9d8776Df7E6E2C4Aad751320eE26164 (Sepolia)
Stack: Solidity · Hardhat · Ethers.js · HTML · CSS · JavaScript
My very first Python project. Built in 12th grade with two friends.
A single-elimination bracket tournament simulator in pure Python — before I knew about frameworks, APIs, or any best practices. Raw, real, and where the journey began.
"A lot of extra code and inefficiency is there... I completed this in my 12th grade with the help of my 2 friends Vishnunath and Dhishaj."
Stack: Python (pure) · Single Elimination Bracket Logic
12th Grade --> Python knockout bracket · First ever project · Zero experience
|
v
Web Dev --> HTML · CSS · JavaScript · how the web works
|
v
Blockchain --> Solidity · Hardhat · Ethers.js · AgriChain on Sepolia ETH
|
v
Full Stack --> React + Node.js + MongoDB · APIs · Authentication · PWA
|
v
AI/ML --> XGBoost · LightGBM · FastAPI · R² 0.9874 on KAVACH DIT model
|
v
Production --> KAVACH live on Vercel · AgriChain live on Sepolia testnet · Now
"Started with a knockout bracket in 12th grade. Now building ML-powered insurance on the blockchain. The journey is just getting started."
If any of my projects helped you, a star means a lot!