Things I’ve built and shipped

Public code, private tooling, production infrastructure, and sanitized work writeups. Each card tells you what can be opened, read, or watched.

Proofmark — Local AI Image Detection

Public source

Proofmark — Local AI Image Detection

A privacy-first Chrome extension that scores and labels likely AI-generated images directly in the browser. The quantized vision model, metadata checks, and evidence UI run locally with WebAssembly—no uploads, backend, or API key.

JavaScriptChrome MV3Transformers.jsONNX Runtime WebWebAssembly
Dino's Essentials — Self-Hosted Platform

Public source

Dino's Essentials — Self-Hosted Platform

A small production platform for the tools I actually want to use. Multiple Dockerized applications run behind a reverse proxy on my VPS, while a deployment daemon watches their GitHub repositories and rebuilds changed services automatically.

Ubuntu ServerDocker ComposeGitHubCaddyNode.jsPython
Drone Trainer — 3D Interception RL

Public source

Drone Trainer — 3D Interception RL

A custom Gymnasium environment for training a pursuer drone to intercept a moving target in 3D. It includes camera-like observations, target-lock memory, fly-through collision geometry, shaped rewards, PPO training, and a lightweight renderer.

PythonGymnasiumPPOStable-Baselines3PygamePytest
LARPmaxxing — Agent-Stack Audit

Public source

LARPmaxxing — Agent-Stack Audit

A deliberately unserious vision app that audits screenshots of coding setups, scores the visible agent and tool sprawl against a structured rubric, and publishes shareable results to a public leaderboard.

Node.jsOpenRouterVision ModelsStructured OutputDockerCaddy
Automated Product Demo Pipeline

Private build

Automated Product Demo Pipeline

A private tooling pipeline that turns repeatable browser interactions into composed product-demo videos. I built it to spend less time manually laying out and recording every launch clip—and more time improving the products themselves.

Node.jsBrowser AutomationVideo CaptureRendering
Kofax Clear — Bug Correlation Automation

Work project

Kofax Clear — Bug Correlation Automation

A human-in-the-loop support automation that combines deterministic browser work with an LLM-powered Azure DevOps search. Selenium extracts failure context; Claude Code reasons over candidate bugs; a strict contract returns parseable JSON.

PythonSeleniumClaude CodeMCPAzure DevOpsJSON Schema
RiverGuard - AI-Powered Trash Detection

RiverGuard - AI-Powered Trash Detection

RiverGuard is a comprehensive video analysis platform that uses machine learning to detect garbage and debris in water bodies. The system processes video streams through a YOLO-based detection model and stores results in Firebase Firestore for tracking and analysis. help clean up rivers and streams.

Next.JsTypeScriptFirebaseCSSPythonYoloV12 MiniDockerUbuntu ServerNGINXVercelFireStore
BabbleBridge - Message Anybody Anywhere, Anytime, Any Language

BabbleBridge - Message Anybody Anywhere, Anytime, Any Language

Babble Bridge is a real-time multilingual chat application that enables seamless communication between users speaking different languages. The app automatically translates messages into each participant's native language, making cross-language communication effortless.

React NativeExpoFirebaseFireAuthFireStoreGoogle Cloud Translation API
Autonomous Tracking Camera - AI-Powered Detection

Autonomous Tracking Camera - AI-Powered Detection

Autonomous Tracking Camera is a project designed as a proof of concept to build a camera that can track humans and record videos. It's purpose was to allow me to record videos of myself climbing, and then use AI to track my movements and record the videos.

RPI 4BPythonOpenCVYoloV3Web Sockets
CPU Scheduler Implementation

CPU Scheduler Implementation

This project implements two CPU scheduling algorithms:Shortest Remaining Time First (SRTF) and Highest Response Ratio Next (HRRN).

Python
Zero Knowledge Chess - Zero Knowledge Proofs for Chess

Zero Knowledge Chess - Zero Knowledge Proofs for Chess

A chess game implementation in Rust that uses Zero-Knowledge proofs to verify the validity of chess games.

RustZero-Knowledge Proofs
Dino's Finance Essential

Public source

Dino's Finance Essential

A self-hosted receipt finance tracker that uses OCR to extract spending data from uploaded receipts, then organizes it with a Next.js frontend, FastAPI services, PostgreSQL persistence, and authenticated user sessions.

Next.jsPythonFastAPIPostgreSQLDockerOCR
SpeedReader - RSVP-Based Speed Reading App

Public source

SpeedReader - RSVP-Based Speed Reading App

SpeedReader is a local-first speed reading app that uses Rapid Serial Visual Presentation to flash words at a configurable pace with optimal recognition point highlighting. It supports browser-based reading workflows for pasted text, uploaded documents, progress tracking, and privacy-conscious client-side processing.

Next.jsTypeScriptReactZustandpdf.jsJSZipCSS