Selected Work

Projects that show engineering depth, product judgment, and range.

Applied AI + Product Systems

Jumpshot Trainer

A two-part biomechanics analysis system for basketball shooting form. I built a Python ingestion pipeline for raw clip processing and a Rust desktop app for shot stage detection, diagnostics, review, and training-readiness analysis.

Python
Rust
MediaPipe
YOLOv8
Parquet

OutcomeTransforms raw video into actionable shooting diagnostics.

StackPython • Rust • MediaPipe • YOLOv8

Hard Part Turning raw video into structured shot-stage diagnostics

System Shape Python ingestion pipeline plus Rust analysis and desktop review

Why It Matters It bridges messy real-world input and a productized feedback loop

shot stage timeline ready -> load -> set point -> release

Release Timing 178 ms

Elbow Alignment +4.2 deg

Readiness Score 81 / 100

Corpus Quality Training Ready

Operations SaaS

Tadester Ops / TadesterFlow

View GitHub repository

A field operations platform spanning a Next.js marketing site, a TypeScript backend API, Supabase data model, and a Flutter mobile app with role-aware dashboards, worker routing, assignments, and tracking.

View Live Landing Page

Frontend Next.js marketing and auth flows

Backend Typed API routes, routing, geofence, worker logic

Mobile Role-aware Flutter workspace for admins and field crews

OutcomeUnifies dispatch, routing, and field execution in one product.

StackNext.js • TypeScript • Supabase • Flutter

> nextjs landing page + waitlist funnel online

> node/typescript api with routing + geofence services

> flutter workspace for admins, dispatchers, and workers

Delivery Operations

Bakery Delivery

View GitHub repository

A Flutter + Supabase operations app for daily bakery distribution, with admin tooling for delivery plans, route assignment, and monitoring plus a driver flow for shift tracking, capacity-aware loading, optimized routes, and delivery confirmation. It is already used in the real world by a bakery team of 10 employees.

Flutter
Dart
Supabase
PostgreSQL
Google Maps

OutcomeUsed by a real bakery team of 10 employees.

StackFlutter • Dart • Supabase • Google Maps

Hard Part Coordinating route assignment, load batches, and real delivery state

System Shape Role-aware mobile app backed by Supabase auth, data, and routing logic

Why It Matters It turns a messy physical workflow into a usable operations product

Project Directory

Everything currently in the workshop.

Applied AI

Jumpshot Trainer

GitHub

Operations SaaS

Tadester Ops / TadesterFlow

GitHub

Delivery Operations

Bakery Delivery

GitHub

Simulation

WorldSim

GitHub

AI Learning Tool

Study Application

GitHub

Portfolio

Portfolio Website v2

GitHub

Visual Gallery

Product visuals across Tadester Ops, TadesterFlow mobile, and Jumpshot Trainer.

Bakery Delivery

Live mobile operations screens

Tadester Ops

Current live visuals

Jumpshot Trainer

Biomechanics review gallery

WorldSim app

WorldSim app gallery

Build Breakdown

What I built, how I built it, and the technologies behind each project.

Jumpshot Trainer

Applied AI biomechanics platform

What I built: a Python video-ingestion pipeline and a Rust desktop analysis app for basketball shooting diagnostics.
How I built it: raw clips are processed into structured data, segmented into shots, scored through stage-based logic, and surfaced in a desktop review experience.
Languages and tools: Python, Rust, MediaPipe, YOLOv8, OpenCV, Parquet, and custom state-machine logic.

Tadester Ops

Field operations SaaS platform

What I built: a full product stack with a marketing site, backend API, Supabase data model, and a Flutter app for admins, dispatchers, and field workers.
How I built it: the system combines auth, organizations, jobs, assignments, routing, worker tracking, and role-aware dashboards across web and mobile.
Languages and tools: TypeScript, Node.js, Next.js, Flutter, Supabase, PostgreSQL, Google Maps services, and Zod-based validation.

Bakery Delivery

Mobile delivery operations platform

What I built: a role-based mobile app for admins and drivers covering delivery planning, assignment, route execution, and completion tracking.
How I built it: the system combines delivery plans, stop management, route assignment, capacity-aware load batches, Google Maps routing, and live status monitoring.
Languages and tools: Flutter, Dart, Riverpod, GoRouter, Supabase Auth, PostgreSQL, Row Level Security, Google Directions API, and Google Places API.

WorldSim

Simulation sandbox built in Rust + Bevy

What I built: a simulation prototype with modular systems for agents, life cycles, world resources, magic, and UI dashboards.
How I built it: the app is structured as Bevy plugins that separate simulation flow, world state, agent logic, and on-screen controls.
Languages and tools: Rust, Bevy, ECS patterns, modular game systems, simulation clocks, and dashboard UI.