Monitors user engagement through computer vision to detect productivity levels in real time.

Chrome extension gauges browsing activity’s relevance to the user’s current task using AI-based classification.

Provides a shared environment for friend groups to raise a virtual duck, adding social accountability.

Triggers humorous AI messages to the group chat if a user’s productivity falls below a set threshold.

Optionally deducts the user’s money or sends automated emails to bosses/family when productivity is low.

Predictive LSTM model that forecasts future spending based on real transaction data.

Interactive 'Oracle' engine where users adjust spending parameters to see impact instantly.

Stunning animated UI built with Tailwind CSS, Shadcn, Three.js, and Next.js.

Secure backend using Convex DB with built-in encryption and Plaid integration for financial data.

Anomaly detection to flag unusual transaction patterns in user history and send real time emails.

Custom dashboard with dynamic charts, category breakdowns, and a Gemini wrapper for smart suggestions.

A sketch model trained with Pytorch and Sckit-Learn that decodes user drawn sketches.

A text model that uses Levenshtein, Fuzzy, and SpaCy to match metadata with user descriptions.

Beautifully crafted UI with Shadcn UI, Tailwind CSS, and Next.js.

Robust authentication and data flow with Clerk, Prisma, and Supabase.

Built a React JS interface with Bootstrap for smooth user experiences, secured by Auth0.

Used MongoDB to store image data and track user activity for leak forensics.

Leveraged Databricks and pretrained AI models for deepfake detection and image processing.

Provided post, view, and download metrics to identify potential media leak sources.

Implemented a cross-platform user interface with React Native.

Utilized SQLite, MMKV, and Filesystem for robust local storage and offline support.

Employed Git and Expo for version control and deployment.

Published the final product on the Google Play Store with 100+ downloads.

Developed Python algorithms for automated crypto trading on Robinhood.

Implemented encryption to secure API credentials and transaction data.

Executed real-time trades and tested performance using backtesting techniques.

Developed a React JS front-end for an engaging user experience.

Combined MongoDB (for AI) and Flask (for RAG + LLM) to summarize NASA's database.

Implemented secure backend systems with Supabase and Express.js/Node.js.

Created original pixel art and assets via Stable Diffusion and Adobe tools.

Built a teacher/student flow with robust authentication for gamified learning.

Led a 5-person team over two days, presenting to judges at NASA Chicago.

Utilized Python to automate server tasks and manage user interactions.

Implemented a simulated stock market system for buying and selling stocks.

Fetched random posts via a Reddit command for fresh content.

Created a customizable text encryption system that allowed users to create their own secret languages.