SPSteven PierceSoftware engineer

Junior software engineer

Software for the
real world.

I build and operate full-stack, mobile, cloud, and computer-vision products—from the camera on the runway to the application in a customer's hand.

Explore projects View résumé
03Production case studies
30K+ framesComputer-vision data pipeline
2 serversSelf-hosted production failover
End to endProduct, code, releases, operations

Selected work

Projects that ship.

Each project began with a real operational problem and grew into a working product with users, infrastructure, and release discipline.

Authenticated PV Local Labeling workspace reviewing a 94 percent run-up prediction and detected athlete02
Computer vision + live operationsLive private operational system

PV Video Capture

A multi-camera capture and clip-delivery system with two custom computer-vision models and more than 30,000 labeling-ready frames.

PythonOpenCVUltralyticsMediaPipeFFmpeg
View project page Live app
Authenticated Landon Pole Rental inventory browser showing searchable pole groups03
Cross-platform commerceLive production system

Landon Pole Rental

An iOS, Android, and web rental platform for inventory, payments, agreements, staff operations, and time-sensitive waitlists.

TypeScriptExpo RouterReact NativeCloudflare WorkersD1
View project page Live app

Production ownership

I built the servers, too.

I provision and operate a two-server Linux platform for these public applications. The work includes secure remote access, system services, deployments, Cloudflare tunnels and load balancing, health monitoring, backups, maintenance, and tested failover—not only application code.

  • Systemd services and timers for apps, deploys, monitoring, and maintenance
  • Encrypted USB and NAS backups, retention checks, and restore drills
  • External uptime checks, synthetic user flows, incident evidence, and UPS-aware shutdown
Meet ManagerLandon Check-InPole Rental
Public edgeCloudflare Tunnel + Load Balancing

Host routing, aggregate health, session affinity, and automatic origin failover

PRIMARYLinux server 01

Production apps, health services, backups, and guarded maintenance

FAILOVERLinux server 02

Mirrored services, independent tunnel origin, and recovery capacity

MONITORBACK UPFAIL OVERRECOVER

How I work

AI speed, engineering judgment.

I use coding agents to move faster through implementation, debugging, testing, and documentation—while keeping product intent, system safety, and verification explicit.

  1. 01
    Define the operating problem

    Start with users, constraints, failure modes, and a testable outcome.

  2. 02
    Build in verified slices

    Use Codex, Claude Code, and Cursor with scoped tasks and reviewable changes.

  3. 03
    Run the whole system

    Test, release, monitor, document, and improve against live behavior.