Selected Work

Where the work meets the stakes.

Six engagements across consumer wellness, health-tech, defense, and platform delivery. The thread is the same: AI-ready architecture, executive translation, and governance that holds under pressure.

Book a discovery call Your on-demand AI officer

A row of leather-bound case ledgers with gold lettering on a dark shelf

Consumer wellness / Fractional CTO / AI in marketing

NuCalm rebuild

Rebuilt a working company's web experience beside the live one, without touching it.

NuCalm has a premium product and a business that works. The job was to scale it and bring AI into how it markets, without putting one live sale at risk. I serve as fractional CTO, and I am an investor in the company.

Approach

Audited the live site and every repository first: page map, consistency, orphaned pages, security, and wording.
Set written rules on what the company can and cannot claim, so no AI draft could drift past them.
Built the new experience as a private beta beside the live site. The live site was reference only and was never edited.
Replaced planning assumptions with measured data from the company's own warehouse.
Ran many AI models in parallel for independent reads, with one adjudicator and a person making the final call.
Automated routine staff email and connected the CRM to billing, so follow-ups stopped depending on memory.

Outcome

A complete rebuilt site in front of the executive team in under three weeks.
A storyboard-controlled hero film that can be cut into ad creative for any channel.
No downtime and no change to the live site while the work was done.
Leadership decides what ships. The work is theirs to adopt in whole, in part, or not at all.

Why this matters for your AI bet

This is the pattern for any profitable business: build beside what works, prove it, then switch over on the owner's say.

Health-tech / Operator / CTO

Transcend turnaround

Stabilized a health-tech platform and restored engineering velocity.

Brought in as senior technology leader during a period of platform instability and team churn. The product served clinical and operational workflows where downtime is not abstract.

Approach

Audited architecture, vendors, and on-call posture in the first weeks. Cut noise before adding scope.
Re-sequenced the roadmap around revenue-bearing workflows and regulatory exposure.
Rebuilt engineering rituals so the team could ship without heroics.
Translated between clinical operators, executives, and engineering so decisions stopped stalling.

Outcome

A platform that operators trust on a normal Tuesday, not just during a launch.
A delivery cadence the executive team could plan against.
A team retained and refocused on the next phase of the business.

Why this matters for your AI bet

If your AI bet sits on top of a shaky platform, the AI is not the first problem to solve.

Defense / Intelligence / Delivery lead

Ukraine Intelbook

Shipped an intelligence product under wartime constraints.

Operational tempo did not allow for traditional acquisition timelines. Multiple agencies, multiple classification boundaries, real consequences for delay.

Approach

Reduced the surface area of the first release to what users would actually open on day one.
Coordinated across agencies and partners without letting governance become a stall tactic.
Built a feedback loop with the people closest to the mission so the product moved with the situation.

Outcome

A product fielded inside an operational window that traditional programs would have missed.
A pattern other teams could reuse for time-critical delivery.

Why this matters for your AI bet

Speed without governance is reckless. Governance without speed is irrelevant. The work is in holding both.

Defense / Theater systems / Executive translator

EUCOM theater systems

Modernized systems across a combatant command without breaking what worked.

Legacy stacks, allied partners, and operators who measure systems by whether they hold up at three in the morning. Modernization had to respect what the field already trusted.

Approach

Mapped the real dependencies, not the org chart version.
Sequenced changes so each release reduced risk before it added capability.
Translated between operators, engineers, and senior leadership so the same decision was understood the same way in every room.

Outcome

A modernization path the field accepted instead of resisted.
Clearer ownership across teams that had been quietly duplicating work.

Why this matters for your AI bet

Most AI rollouts fail at the translation layer, not the model layer. This is that work.

Defense / Classified / Program delivery

NRO program delivery

Delivered in a classified environment with the governance to match.

High-sensitivity program where the cost of a mistake is not measured in churn. Risk, oversight, and pace had to coexist.

Approach

Built a risk posture that leadership could defend on the record.
Kept the team moving by making the governance legible instead of ceremonial.
Closed the loop between technical decisions and program-level accountability.

Outcome

A program that passed scrutiny without losing its delivery rhythm.
A model for governance that engineers stopped resenting.

Why this matters for your AI bet

AI governance, model risk, and audit posture are the same muscle, applied to a new domain.

Health-tech / 503A pharmacy / Platform

503A pharmacy platform

Took a HIPAA-conscious pharmacy platform from pilot to production.

A 503A workflow touches prescribers, pharmacists, fulfillment, and patients. Each handoff is a place errors compound. The platform had to be defensible to clinicians, auditors, and operators at the same time.

Approach

Designed the data model around regulatory reality, not convenience.
Wired observability into the clinical workflow so failures surfaced before patients felt them.
Made the build-versus-buy calls explicit so leadership could fund them with eyes open.

Outcome

A platform clinicians and operators rely on for daily work.
An infrastructure base that later AI features could sit on without rework.

Why this matters for your AI bet

Regulated AI lives or dies on the data and workflow plumbing underneath it. That plumbing is the work.

The thread

AI-ready architecture, executive translation, governance under pressure.

Different sectors. Same job. The work is making sure the technology, the people, and the accountability move at the same speed.

See how this applies to your roadmap.

One conversation. No deck required. We will know quickly whether there is a fit.

Book a discovery call AI consulting overview