I'm a third-year CSE (AI & ML) student at UPES Dehradun who builds things end-to-end — from edge AI on ESP32 and Raspberry Pi to NLP-powered web platforms to NASA Earth data apps. I've led teams at two of India's biggest hackathons and shipped real products under pressure. I care about clean architecture, useful AI, and code that actually solves something.
A personalised weather app that goes beyond standard forecasts. Uses the ACIS (Activity-Centric Impact Score) — a custom 1–10 risk metric — to tell you whether a day is a 9/10 for cycling or a 3/10 for a picnic, and why. Includes an Optimal Day Finder that automatically identifies your best window across the forecast range.
What makes it different: Standard weather apps give you temperature and precipitation. ChronoClime fuses temperature, wind, humidity, UV, and AQI into a single activity-specific score, then surfaces the best day automatically. Includes an optional ESP8266 IoT hardware gateway for GPS-based location input and an AI weather chatbot for context-aware queries.
My contributions: Designed and implemented the ACIS scoring algorithm (the core differentiator), built the Activity Planner component, integrated the OpenWeatherMap API pipeline, developed the ESP8266 IoT gateway integration, and built the AI chatbot service.
React TypeScript Vite Tailwind CSS Framer Motion Supabase Leaflet ESP8266 Arduino FastAPI
Built for Indian Railways. An AI-powered lifecycle traceability system that assigns every track component a unique laser-etched QR identity. Served as Software Lead, owning the full AI ecosystem.
What I built: React frontend for field engineers to scan QR-marked components and retrieve complete maintenance history; AI analytics dashboard (Python/Dash/Plotly) with failure-rate visualisations, component lifespan forecasting by zone, and warranty alerts; cloud backend (Firebase/AWS) with REST APIs syncing bidirectionally to Indian Railways' UDM and TMS portals.
Raspberry Pi Coral Edge TPU MobileNetV3 React Firebase AWS Python Dash Plotly REST APIs
A web-based AI platform that takes a hackathon problem statement, maps it to the relevant UN Sustainable Development Goals using NLP, and outputs structured project ideas with a full week-by-week execution roadmap.
Architecture: Single-file React frontend → FastAPI REST API (6 endpoints) → Scikit-learn NLP pipeline (TF-IDF + Logistic Regression) + NLTK preprocessing · PostgreSQL + MongoDB data layer
Python FastAPI Scikit-learn NLTK TF-IDF React PostgreSQL MongoDB NLP REST APIs
A Tinder-style mobile-first shopping app — swipe right to wishlist, swipe up to cart, swipe left to skip. Built with React + TypeScript + Capacitor for native iOS deployment.
My contributions: Built the ProductCard swipe engine — touch and mouse gesture detection, directional swipe logic, overlay animations, and double-tap support. Developed CategoryTabs and real-time search filtering. Handled all UI polish across components.
React TypeScript Vite shadcn/ui Tailwind CSS Capacitor React Query iOS
Edge AI system for predicting machinery failures before they occur — without needing historical failure data. ESP32 runs hardware-accelerated FFT locally, reducing network payload by over 90%. Isolation Forest model learns the healthy baseline and flags anomalies without labelled data. React Three Fiber 3D Digital Twin dashboard shows live machine health.
My contributions: Built the React frontend and 3D Digital Twin dashboard, developed the FastAPI backend with SSE streaming for live data push, assisted with the CSV data pipeline.
ESP32 Arduino FFT Python FastAPI Scikit-learn Isolation Forest React React Three Fiber WebGL Recharts
Online exam system with JWT auth, service discovery, and an API gateway built on Spring Cloud. 5 independently deployable services communicating via REST.
Spring Boot Spring Cloud Eureka JWT MySQL React
Expense tracking, live market updates, and investment learning modules — built to make financial literacy accessible to young users.
Python Financial APIs Data Analytics
- 🏆 Grand Finalist — NASA Space Apps Challenge 2025 ("Will It Rain On My Parade?" — ChronoClime)
- 🏆 Grand Finalist — Smart India Hackathon 2025 (SIH25021 — Team KASA, UPES representative)
- 📌 Minor Project — IDEACT, AI-powered SDG hackathon advisor, UPES Dept. of CSE (2025–26)
- 💼 Internship — Binary Keeda Education Pvt. Ltd. · drove 50% audience engagement growth in 2 months (Jun–Aug 2025)
- 💼 Internship — Chhanv Foundation · field campaigns and awareness sessions, Sheroes Hangout Cafe (Jun–Jul 2024)
Actively looking for internships & collaborations in backend, AI/ML, or full-stack roles. Also happy to contribute to interesting open-source projects. If you're building something ambitious, let's talk.
