Agentic AI fitness companion. Inspired by my love for fitness, technology, and natural language interfaces. Solo build. Developed the iOS app using React Native and a python backend for authentication, in-app payments, and AI communication. AI layer built atop LangGraph with custom solutions for memory, context management, and evals. Integrated with frontier lab LLMs. Supabase for DB.
Interactive visual illustrating Paul Graham's disagreement hierarchy, with AI-generated examples. Built in Google's AI Studio and hosted via Google Cloud Platform's ecosytem of services.
AI-powered meal planning agent that suggests recipes, scales portions, and generates shopping lists tailored to user preferences and dietary needs. Solo hackathon project. Developed without a framework by defining tools that integrate with recipe database web services and an agentic loop that integrates with OpenAI's chat completions API.
Intuitive watchOS scorecard app that eliminates distractions so users can focus on their round. Grew to 20 monthly active users and 61+ scorecards per month. Led technical design and development, managed a remote development team. iOS + WatchOS app built in Swift and backend built with node.js. MongoDB for noSQL DB.
Design‑led, automated DTC apparel project inspired by New York City's most underrated bridge. I integrated an ecommerce platform with Stripe payments and a drop-ship supplier to automate order placement, fulfillment, and payments processing.
Calculator for Chrome. Scaled to 2,200 weekly active users and earned 4.8-star review. At the time, there was no easy-to-use and intuitive native calculator app for value market Windows and Chrome OS devices. I built an easily accessible calculator with basic HTML, CSS, and Javascript and launched it on the Chrome App Store. I was shocked by the adoption. I learned firsthand the power of the internet's scale for distributing software.