All projects
Featured work across AI/ML systems, graph analytics, regulatory intelligence, and desktop software.
AI / ML Project
SATARK: AI-Powered Disaster Management System
AI-based disaster classification for Floods, Cyclones, and Wildfires using Vision Transformers (ViT), achieving 98.32% cyclone and 94.39% flood detection accuracy. Features a Flask dashboard with real-time disaster visualization and smart resource allocation using the Haversine formula, backed by ML evaluation pipelines with ROC curves and confusion matrix analysis.
Data Science & Graph ML
Professional Network Analytics & Graph Intelligence
Processed and analyzed 2,400+ LinkedIn profiles, extracting and structuring 10,000+ skill, follower, and experience data points using Pandas, Regex, and AST parsing. Built a 2,400+ node network graph in NetworkX computing PageRank, Degree, Betweenness, and Closeness Centrality, producing 20+ automated statistical visualizations validated through automated Pytest pipelines.
Full Stack & Regulatory AI
Krio.LexGov: Public Regulatory Intelligence Platform
Full-stack platform collecting and organizing 39 SEBI enforcement orders into searchable records containing case, entity, penalty, and regulatory details. Built with FastAPI and SQLAlchemy for PDF extraction, data normalization, and SHA-256 source tracking, paired with a Next.js + TypeScript frontend offering regulatory analytics and an AI-assisted precedent search feature.
Desktop & Productivity Engineering
Distrack Desktop
Cross-platform desktop application designed for distraction tracking and productivity management. Built to monitor and analyze workflow patterns, deep work sessions, and context-switching behaviors to help engineers maintain focus and optimize productivity.