Career narrative
A working record, not a role announcement
This site is the public version of a project portfolio compiled in September 2026. It is not a resume written for one job title. It is a record of how I actually build things: architecture first, AI inside explicit boundaries, validation with evidence, and a human decision where being wrong is expensive.
01
Operational security systems
Professional experienceOn the Chelan County PUD service desk I joined a knowledge-publishing effort led by a senior IT specialist. I did not lead that project. Running real Self Help guides through a multi-stage GPT pipeline taught me that AI is something you architect in stages — and that a person still determined every next action.
02
Automation
IndependentTinyClaw took a daily manual gather and turned it into a locally hosted scheduled pipeline. The design problem was not “call a model.” It was idempotence, freshness, and the willingness to skip a run rather than publish a partial report.
03
AI experimentation
IndependentA Gemini Gem constrained to one vehicle — a 2025 Jeep Gladiator Sport S with Max Tow — made the configured-assistant class concrete. Context engineering and constraint definition did more than model choice. High-impact suggestions still route to me.
04
Agentic systems
Independent home-lab / designThe Knowledge Publishing Agent is a specified design, not yet built. The Endpoint Triage Agent is built, isolated, and adversarially tested in a contained Hyper-V lab. That distinction is intentional. I will not collapse a design and a lab build into the same claim.
05
Application architecture
Independent product / platformClearward ships structured Excel and Google Sheets products under software-grade QA. Basaltborne is a production-track publishing platform with Row-Level Security, migrations, API boundaries, and a multi-layer CI suite. Secure platform engineering there is labeled developing, not finished.
06
Secure AI thinking
ThroughlineAcross every project the same controls recur: least privilege, tool allowlists, untrusted input, deterministic validation, audit, safe-stop, rollback, and human approval. They are not slogans. They are the parts that survived contact with real failure modes.
How the work is labeled
Professional experience
Chelan County PUD service desk. Team member on the ServiceNow Self Help publishing workflow. A senior IT specialist led that work.
Independent projects
Jeep domain assistant, TinyClaw, Clearward, Basaltborne.
Lab work
Home-lab Endpoint Triage Agent, including isolation and red-team testing against my own agent.
Self-directed design
Knowledge Publishing Agent — specified, not yet built. Phase 2 of the published roadmap.
Skills as evidence
Not a decorative cloud. Each skill points at the projects that actually used it.
AI
- Prompt engineeringUsed in: Domain-Specific AI Expert · Multi-Stage Knowledge Publishing Workflow · Clearward
- Context engineeringUsed in: Domain-Specific AI Expert
- AI task decompositionUsed in: Multi-Stage Knowledge Publishing Workflow · Clearward
- Agent supervisionUsed in: Clearward · Basaltborne
- Human-in-the-loop systemsUsed in: Home-Lab Endpoint Triage Agent · Clearward · Basaltborne
- Agentic workflow designUsed in: Home-Lab Endpoint Triage Agent · Knowledge Publishing Agent
- Multi-stage LLM workflowsUsed in: Multi-Stage Knowledge Publishing Workflow · TinyClaw
- Agent state modelingUsed in: Knowledge Publishing Agent · Home-Lab Endpoint Triage Agent
- AI maturity classificationUsed in: Multi-Stage Knowledge Publishing Workflow · Domain-Specific AI Expert · Home-Lab Endpoint Triage Agent
- Domain-specific assistantsUsed in: Domain-Specific AI Expert
- Confidence modelingUsed in: Home-Lab Endpoint Triage Agent
Automation
- Scheduled automationUsed in: TinyClaw
Software Engineering
- Systems architectureUsed in: Basaltborne · Home-Lab Endpoint Triage Agent
- Next.js / React / TypeScriptUsed in: Basaltborne
- API designUsed in: Basaltborne
- Schema validation (Zod)Used in: Basaltborne
DevOps / CI/CD
- Git / GitHub / branchingUsed in: Basaltborne
- CI/CDUsed in: Basaltborne · Clearward
Databases
- Database migrationsUsed in: Basaltborne
- PostgreSQL / SupabaseUsed in: Basaltborne
Cybersecurity
- Row-Level SecurityUsed in: Basaltborne
- Authentication & authorizationUsed in: Basaltborne · Clearward
- Least privilegeUsed in: Home-Lab Endpoint Triage Agent · Clearward · Knowledge Publishing Agent
- Secret managementUsed in: Clearward · Basaltborne
- Input validationUsed in: Basaltborne · TinyClaw · Clearward
- Trust boundariesUsed in: Basaltborne · Home-Lab Endpoint Triage Agent
- Fail-closed behaviorUsed in: Basaltborne · TinyClaw
- Prompt-injection resistanceUsed in: Home-Lab Endpoint Triage Agent · Knowledge Publishing Agent
- Tool allowlistsUsed in: Home-Lab Endpoint Triage Agent · Knowledge Publishing Agent
- Audit loggingUsed in: Home-Lab Endpoint Triage Agent · Knowledge Publishing Agent
Testing / Validation
- Adversarial testingUsed in: Home-Lab Endpoint Triage Agent
- Vitest / PlaywrightUsed in: Basaltborne
- Database-security testing (pgTAP)Used in: Basaltborne
- Accessibility testingUsed in: Basaltborne
- Release gates / hashingUsed in: Clearward
Security Systems
- PowerShellUsed in: Basaltborne · Clearward
- Windows / Hyper-V lab administrationUsed in: Home-Lab Endpoint Triage Agent
- Endpoint telemetry / event analysisUsed in: Home-Lab Endpoint Triage Agent
Product
- Excel & Google Sheets systemsUsed in: Clearward
- Product documentation / UXUsed in: Clearward · Basaltborne
These seven projects sit at different scales — a weekend experiment, a home lab, and two platforms built for real customers — but the underlying discipline does not change: define the architecture and the boundaries first, let AI agents do the drafting and the heavy lifting inside those boundaries, validate what comes back with evidence rather than trust, and keep a human decision at every point where the cost of being wrong is real.