Program direction
Set outcomes, align stakeholders, manage delivery risk, and keep complex modernization work moving through real federal constraints.
Federal mission systems / cloud / applied AI
I lead where mission, technology, and delivery meet.
GS-14 program leader doing solutions architecture work inside FDA’s Center for Tobacco Products. I modernize research systems, embed responsible AI into regulatory workflows, and turn high-stakes problems into systems teams can ship and sustain.
20+Years in federal technology
400+Regulatory scientists served
GS-14Program leadership / COR III
15M+Records integrated in CIRDS
01Operating range
Technical enough for the architecture. Senior enough for the institution.
The most consequential federal technology work crosses mission priorities, acquisition, security, data, users, vendors, and operations. I work across that full system—especially for CTP’s regulatory science and research mission.
Technical translation
Move fluently between an executive brief, a procurement decision, a user workflow, and an architecture review.
Product judgment
Find the smallest useful release, make the tradeoffs explicit, ship it, and build evidence for the next decision.
Featured build / 2026
SAA—
Swipe
A focused iPhone study companion for AWS Solutions Architect candidates. It turns architecture tradeoffs into guided lessons, scenario cards, and a clear readiness signal.
This is more than certification prep. It is a product system: content architecture, interaction design, progress modeling, native iOS delivery, and AWS-backed distribution.
SwiftUIAWS architectureProduct designIn active development
PetShots / production topologyServerless by design
CloudFront
+ S3
+ S3
Cognito
JWT
JWT
API Gateway
+ Lambda
+ Lambda
S3 records
+ events
+ events
15API routes
0Idle servers
CDKInfrastructure as code
Production platform / live
PetShots
Secure pet records and revocable QR passports, designed and deployed end to end on AWS. The architecture favors low operational overhead, strong data boundaries, and costs that match an early-stage product.
The interesting work is in the decisions: direct-to-S3 uploads, authorization at the gateway, retained user data, and a deliberate migration path when today’s simple storage model stops winning.
TypeScriptAWS CDKEventBridgeCognito
Applied AI / inside the firewall
AI embedded in FDA research tracking—not bolted on.
I led the integration of an in-boundary LLM into CTP research tracking applications. It analyzes project abstracts against the coding manual, surfaces supporting evidence, and flags discrepancies. Regulatory science reviewers keep the final call.
View the operating model01
Constrain the task
Analyze against an authoritative manual—not an open-ended prompt.
02
Show the evidence
Return supporting snippets and discrepancy flags with each recommendation.
03
Preserve accountability
Use AI to narrow the discussion; keep decision authority with reviewers.
04
Fit the environment
Operate within the agency boundary and the existing review process.
The through-line
Make the complicated workable—and make the work matter.
Air Force veteran. Computer Science graduate. MBA. GS-14 program manager and COR III. I’ve spent two decades seeing how government technology succeeds, stalls, and survives contact with reality. My FDA work spans research data integration, scientific information retrieval, product delivery, and responsible AI.
More about how I operate02Field notes