Projects

These are some of the projects I’ve built over the years. Each one came from a specific need, curiosity, or personal interest rather than just a coding exercise. For more details and code, visit my GitHub profile.

  • AI Call Bot - Hand Coded A Go-based voice-driven AI agent deployed as a Digital Ocean serverless app, accessible via phone demo at (734) 766-7700.
    This project reflects my passion for communication systems, bringing together telephony, AI models, real-time speech recognition, and voice response into a production-style environment.

  • Simple Key Value Server (SKVS) - Hand Coded
    A tiny, key-value store written in Go and deployed over UDP.
    I built this because Redis and similar systems carried far more operational surface than my workload justified. My API needed a fast, read-heavy, single-owner cache with strict latency guarantees, no persistence, and nothing extra. By owning both ends of the protocol, I kept the wire format minimal, encryption mandatory, and behavior predictable.
    View on GitHub → and Read about the design philosophy here

  • SimplyAuto - AI Assisted A free, lightweight auto clicker and macro recorder for Windows, built entirely in Go. Features customizable click intervals, macro recording/playback, global hotkeys, and save/load functionality. Available on the Microsoft Store, or as a portable .exe with no installation required. Visit Site → · Get it on the Microsoft Store → · View on GitHub →

  • wxs.us (Weather Service) - AI Assisted A Go-based weather service hosted on AWS, delivering US forecasts, alerts, and severe weather outlooks via curl, an interactive SSH dashboard, and a native Android app. Built on NWS and NOAA Storm Prediction Center data, it requires no API keys or accounts. Features 7-day forecasts with ASCII art, active weather alerts, severe weather outlooks, and hourly breakdowns. Visit Site → · Get it on Google Play →

  • Writers Grove - AI Assisted A writing platform built with Go, designed for writers who work better with structure. Features world-building tools, character tracking, and organizational workflows that keep complex creative projects manageable. Currently in active development. Visit Site →

  • FriendTime - AI Assisted A web service for coordinating gaming sessions with friends. Finding a time when everyone can actually play is the hardest part of multiplayer gaming, so FriendTime handles the scheduling so the group can focus on the game. The frontend is vanilla TypeScript with a hand-rolled router, no framework and no npm. The backend is Go on a single arm64 Lambda, fully serverless on AWS (S3 + CloudFront, API Gateway, SES) with Turso for the database and passwordless magic-link login. Visit Site →

  • JackBot - Hand Coded
    A Discord-based conversational AI bot written in TypeScript.
    I built it for my kids, naming it after their cat. It became a fun way to experiment with event-driven programming, API integration, and personality-driven design while making something meaningful to my family.
    View on GitHub →

  • GlyphEngine - Hand Coded A small game engine built, using the Ebitengine framework, to power The Social Contract. View on GitHub →

Tools Link to heading

  • TTS (Text-to-Speech CLI) - Hand Coded A command-line tool for converting documents into speech using AI.
    I originally wrote this before mainstream tools made the process simple. It solved my own problem of wanting to listen to written material in a flexible, scriptable way.
    View on GitHub →

  • Project Tracker - AI Assisted A modular terminal dashboard for tracking project status, built with Go and Bubble Tea. Features vim-style navigation and multiple modules including CI status, TODO, notes, and git status. View on GitHub →

  • GHFlow - AI Assisted A terminal dashboard for monitoring GitHub Actions workflows across multiple repositories. Features a 3x2 grid layout, live status updates, and vim navigation. View on GitHub →

  • aws-sec - AI Assisted A thin Go wrapper around the AWS CLI that intercepts responses and replaces secret-shaped values with [REDACTED] before they reach the caller. Built as a hard boundary for LLM-controlled shells, where environment variables and credentials in normal AWS output would otherwise leak into model context, transcripts, and downstream tool calls. View on GitHub →

Sites Link to heading

To see more of my work visit my Github Repository page