WINSTON-SALEM, NC
STATUS: AVAILABLE
TERMINAL X TERMINAL X TERMINAL X
TERMINAL X TERMINAL X TERMINAL X TERMINAL X TERMINAL X TERMINAL X
ISIAHM R. WARDLOW

AI Implementation & Enablement

I build and ship production AI by directing agents — and help teams actually adopt it.

Open to freelance, contract, and full-time roles
See my work Get in touch
SCROLL

// SELECTED_WORK
PRODUCTIONOTW Cleaning
90+ delivery batches · security-audited · deployed
Cleaning-operations SaaS with an AI-powered client portal for scheduling, dispatch, and job management.
FastAPIReactPostgreSQL
PRODUCTIONAt Ease Property Management
50 routes · 319 passing tests · live on DigitalOcean
Rental-management platform with AI maintenance triage for landlords and tenants.
FastAPINext.jsPostgreSQL
PRODUCTIONThe Kingdom & The Keep
660 questions across 132 quests · live on Vercel
Gamified certification-study app that turns exam prep into quests and progression.
Next.jsSupabaseVercel
PRODUCTIONHerald
Tenant-isolated · human-review gates · audit logging
Governed multi-tenant content-publishing system with least-privilege access and security-audited release gates.
Next.jsSupabaseVercel
CLIENTSimply Blessed
Passed 7/7 QA criteria · live
AI customer-service agent built on Anthropic Managed Agents for a client business.
Anthropic Managed Agents
CLIENTLively Stones (“Gary”)
Passed 5/5 tests · live
Compliance-aware AI voice agent for a financial-services client, built to a 15-section production spec.
GoHighLevelTwilio
PRODUCTIONThe Steward
22+ build batches · AI thesis review · rules-based lockout engine
Trading-discipline app with AI coaching and behavioral nudges that help users hold to their own rules.
AI CoachingBehavioral Design

// ABOUT.md

I'm Isiahm R. Wardlow. I'm not a programmer — and that's the point.

I direct AI the way a general contractor directs the trades. A contractor doesn't swing every hammer or run every wire; they hold the plan, set the standard, and make sure the electricians, plumbers, and framers deliver work that meets it.

I do the same with AI. I write the specifications, set the acceptance criteria, review every batch of work, and ship production software that real people use — not prototypes or demos.

The result is a portfolio built faster than a solo developer could manage, without cutting the discipline that makes software trustworthy: automated tests, security review, and documented decisions at every step.

Isiahm R. Wardlow · Terminal X · Winston-Salem, NC

// HOW_I_WORK.sh
01
Spec first
Every build starts with a written specification and clear acceptance criteria — before any code is written. No vague goals.
02
Batch delivery
Work ships in small, reviewable batches. Each one has a defined "done," so progress is visible and nothing sprawls out of control.
03
Human review gate
I review every batch against the spec before it merges. The AI does the building; I own the standard it's held to.
04
QA & security
Automated tests and adversarial code review run before anything reaches a user — bugs and security holes get caught early.
05
Project memory
Every decision, change, and architecture note is documented, so the work stays maintainable and auditable long after launch.
06
Ship & iterate
Deploy, watch it run in the real world, and improve from there. Working software in production beats a perfect plan on paper.

// AI_ADOPTION
I help teams catch AI failures before they ship.

Most “AI adoption” gets measured by access — seats filled, logins counted, a tool rolled out. I measure it by something harder: whether people can catch a planted failure when one shows up in their work. That’s behavior change, not just access — and it’s the line between a team that uses AI and a team that can trust it.

I build and teach against the failure modes that actually cause damage in production: hallucinated output, prompt injection, and silent errors that slip through unnoticed until they’re expensive. The point isn’t to fear the tools — it’s to build the reflexes that keep a human in the loop at the moments that matter.

Underneath this is a framework I authored — the Seven Laws of the Built World — a map of AI failure modes that I developed, adversarially stress-tested against four different AI models, and am currently piloting as curriculum for teams adopting AI.

// BY_THE_NUMBERS
12+
Platforms shipped
90+
Delivery batches
730+
Automated tests passing
0
Lines hand-coded — every line directed through AI
4
AI models the framework was stress-tested against

// FAQ.dat
Are you available for full-time roles, or only contract?
Both. I take freelance and contract projects, and I'm open to the right full-time role. If you're hiring or scoping a project, the fastest path is the form below.
You say you don't write code — what does that mean for the work?
It means I direct AI coding agents to build the software while I own the specification, the standards, and the review. The output is real, deployed, tested software — think general contractor, not hobbyist.
What does your process look like?
Spec first, then small reviewable batches — each with clear acceptance criteria, automated tests, and a security/QA pass before anything ships. Every decision is documented so the work stays maintainable.
How do you keep AI-built software safe and reliable?
Guardrails at every step: human-review gates before anything merges, least-privilege access, audit logging, and automated tests that must pass before shipping. The AI does the building — the standards, the safeguards, and the final sign-off are mine.
Where are you based, and do you work remotely?
Winston-Salem, North Carolina. I work with clients and teams anywhere, remotely.

// LET'S_TALK

Freelance, contract, or full-time — tell me what you're working on and I'll get back to you.

▷ MESSAGE RECEIVED ◁
Thanks — I'll be in touch soon.