Developed a full-stack legal data processing application using Python (FastAPI, PostgreSQL) and React/TypeScript, implementing automated PDF parsing and statute ingestion from multiple state sources (CA, CO, FL, NV, NY, OR, TX) to handle over 10,000+ legal documents.

Built an end-to-end data pipeline with web scraping, semantic search, and API endpoints, reducing manual data processing time by integrating LLM-based extraction and tagging for accurate legal factor identification in hackathon scenarios.

Led and mentored a team of students through an end-to-end AI safety project, setting direction, milestones, and coordination to deliver a research-grade prompt robustness evaluation system.

Defined project scope and success criteria, facilitated weekly reviews, and unblocked technical challenges while aligning team progress with program stakeholders.

Built a mobile app that identifies plants and analyses their health using AI — combining PlantNet and Google Gemini to deliver real-time care tips, revival steps, and toxicity warnings from a single photo.

Open-sourced on GitHub with React Native/Expo, featuring camera integration, local plant storage, and a configurable setup using free API keys.

Built a Windows desktop utility that summons a circular app launcher around the cursor on a global hotkey, letting users launch any app in one click without ever touching the taskbar or Start menu.

Runs silently in the system tray with zero taskbar footprint; fully configurable via a settings UI for all 8 slots and the activation hotkey, with persistent JSON settings and a file-based crash logger.

Brock Interactive Training Engineering System (BITES) is a comprehensive software engineering training platform designed to enhance user engagement and streamline learning.

Developed using Svelte, TypeScript, Git, and Firebase, the application focuses on delivering an interactive and intuitive training experience.

The project leverages Python to implement a genetic algorithm for cryptanalysis, enhancing decryption efficiency by 20%.

It explores key concepts like fitness evaluation, mutation, and crossover methods, showcasing the potential of evolutionary algorithms in solving complex problems.

An interactive platform developed using HTML, CSS, and JavaScript, allowing users to build custom computers seamlessly.

Designed with responsive design principles, it provides a user-friendly interface and intuitive functionality, combining technical precision with visual appeal.

Features real-time validation to ensure compatibility and optimize user customization.

Developed a AI benchmarking platform evaluating multiple AI agents across games and capability tests, implementing performance tracking, and human-centered evaluation frameworks.

Built a scalable Flask web application with neural networks, API integration, and automated performance analytics for AI models.