Case studies

AI applications, data systems, and operating controls.

Explore the problems I took on, what I designed and built, and what changed as a result. These six case studies cover AI applications, business research, publishing, and production automation. The Home Automation study is coming soon.

Training, recovery, and race analysis

Apex Human Hybrid

CoachAI → Apex Human Hybrid

I evolved CoachAI into a training platform with structured workout history, recovery calculations, route analysis, and web and mobile clients. PostgreSQL/PostGIS and FastAPI support coaching decisions and longitudinal performance tracking.

Working private application PostgreSQL/PostGIS · FastAPI · Vue/React

Read the case study

Job discovery through application review

Job Hunt OS

I built a workflow that collects job opportunities, scores fit, and prepares application packets for my review. The September 3, 2026 snapshot contains 40,081 canonical job records. Every application is submitted manually.

Working internal application Python · Apps Script · Playwright

Read the case study

Learning missions and earned arcade time

Level-Up Learning

I built a privately hosted learning application shaped by my son's feedback, with 3,158 playable missions and four reusable game engines. An offline AI pipeline prepares content, while practice sessions, mastery tracking, and arcade rewards run in the browser.

Privately hosted application React · Node.js · JSON Schema

Read the case study

Public records into useful business research

New Pencil Lead Intelligence

I built a pipeline that turns Texas permit records into business research and assets ready for review. It preserves source history and human notes in Google Sheets, with 6,100+ normalized records and 98.7% coordinate coverage in the local artifact set.

Working core data pipeline Node.js · Google Workspace · Sharp

Read the case study

Operating and modernizing WordPress Multisite

WordPress Modernization & AI Operations

I operate and modernize a production WordPress Multisite platform with checksum-gated changes, custom OIDC identity, focused MU plugins, and offline AI content validation. Fourteen tests check claims, dates, and affiliate markup before editorial review.

Verified production operations WordPress · PHP · Python

Read the case study

Mac reliability, backups, and recovery

macOS Maintenance & Hardware Revival

I built maintenance and recovery tools across household Macs, network storage, and container services. The system combines verified file transfers, PostgreSQL backup mirroring, restore checks, and retention rules that preserve older backups until recovery is proven.

Deterministic system Python · zsh · PostgreSQL · Docker

Read the case study

Full-stack home automation

Home Automation

A full-stack Home Automation case study is coming soon. It will cover the shared container platform, integrations, automation rules, state, recovery, privacy boundaries, and the reasons the system evolved.

Coming soon Home automation · shared containers

View coming-soon note

Evidence language

How project evidence is classified.

Verified evidence: Backed by reviewed code, prompt and knowledge files, tests, dated operating evidence, or active use. Each project page names the verified component and current limits.
Owner-confirmed: Confirmed by Brett without complete component-level repository or operating evidence.
No AI: Conventional software handles this reliability workload.

The labels keep verified evidence, owner-confirmed context, and conventional software with no AI runtime distinct.