Executive Summary
Decade-plus track record building, scaling, and modernizing technical organizations across software engineering, data, AI, and enterprise IT. Current portfolio: $1.2M in realized annual value across 25 production systems, measured bottom-up across nine value channels (revenue protected, retention, error prevention, billing accuracy, and more), with every dollar a live formula against auditable inputs. Add 176 staff hours returned weekly (≈4.4 FTE) and a further $200K+ built and demonstrable, awaiting adoption.
I manage engineering, enterprise IT, and the AI program as one function. I have built and led full-stack engineering teams, internal IT and red-team functions, project-management and asset organizations, and a non-profit wing of a major cancer-care organization. I pair a clear strategic vision with deep technical credibility and disciplined change management, and I move organizations from talking about transformation to finishing it.
What I Run
Engineering
A tight team of engineers who own their systems end to end, from spec to on-call. 25 systems in production, each instrumented, supported, and adopted by frontline teams. Nothing is worth a dollar in the model until it runs for real users.
AI, end to end
Strategy, policy, vendor selection, deployment, training. Chose and rolled out Claude Enterprise and OpenAI Enterprise for 100+ staff; Claude on AWS Bedrock in production agents; local models (CLIP, Ollama) where data cannot leave the building. Chair of the AI and Automation committees.
Enterprise IT and security
IT for 100+ employees: IAM hardening across AWS, GCP, and Google Workspace, hardware tokens, quarterly security training, vendor assessment rubric, 24-hour incident protocol.
Innovation with a price tag
A nine-channel value model behind every build, with COGS, margin, and EBITDA impact modeled per system. Headline numbers get audited, and revised downward when they should be.
Workflow automation
176 staff hours a week handed back. A daily compliance sweep across roughly 1,400 ad accounts, a four-hour reconciliation loop across eight platforms, and a build-vs-buy that retired a $50K/yr vendor feature.
Client-driven, revenue-bound
Two productized offerings shipped, an internal tool turned into a metered billable service, $250K+ in technical consulting sold, and a proprietary dataset that compounds in value as it accrues.
Selected Outcomes
$1.2M realized annual value, modeled like a P&L
Bottom-up valuation across nine channels: revenue protected, client retention, media efficiency, error prevention, billing accuracy, risk avoided, labour returned, cost avoided, and new revenue. Every figure is a live formula with a confidence rating. No asserted totals. The latest audit revised the headline downward on purpose; defensible beats big.
25 systems in production
AI agents, automation engines, Chrome extensions, internal web apps, and self-serve platforms, all instrumented, supported, and adopted by frontline teams. Several have matured into standalone IP, including a billable productized service and a proprietary dataset that compounds in value.
Authored enterprise AI Policy v2.0
Co-authored the org-wide AI Guidelines: 7 core principles, 16-category Universal Data Restrictions, 3-tier tool tiering, vendor assessment (SOC 2 / ISO 27001 / breach notification / training-data opt-out / subprocessor transparency), 24-hour incident protocol.
Two productized offerings shipped
New product lines crossed $50K revenue inside their first six months; two more applications planned for standalone release.
120% revenue lift & 570% optimization yield
Architected the proprietary data and reporting platform that saved the firm millions in 2022.
$250K+ technical consulting sold
Sold directly in a sales-engineer capacity while building the data lake that retired $90K/yr in warehousing costs.
Innovation Highlights
Agentic AI ad-buying agent (first-in-class). LLM-driven media buyer on Claude via AWS Bedrock with 36 custom platform tools. Takes raw emails and PM input and executes buys across DSPs and social platforms with full PM-system traceability. Shipped before agentic media buying went mainstream.
Pixel Lab. Chrome extension for safely testing ad pixels and proving consent compliance: live view of every platform collecting data, consent simulation with plain-English proof of what each platform learned, ad-click-to-purchase journey checks, and a Safe mode that blocks delivery at the browser level. In private beta.
Bid Inspector. Chrome extension that shows the header-bidding auction as it happens in the browser: every Prebid, Amazon TAM, and Google Ad Manager bid including the ones that lost, the creatives and advertisers behind them, decoded identifiers, consent state, and the targeting sent. Includes a targeting injector with per-rule verification, session snapshots with compare, and redacted HTML and Excel reports. First of its kind. In private beta.
Creative Finder. Semantic search across a library of thousands of mail pieces and digital ads. Plain-English queries ranked by a CLIP model running locally, so no creative leaves the network. Facet filters, PDF rendering, ZIP and slide-deck export, usage monitoring. Flask and SQLite, deployable on a LAN Mac or a $22/month cloud box.
Datum Veil. Privacy-preserving Python tool combining HMAC pseudonymization, sum-preserving perturbation (three modes), and Paillier homomorphic encryption to enable safe use of proprietary datasets with third-party AI systems. Authored leadership-facing documentation translating cryptographic guarantees into business terms.
AONS, the Ad Object Naming System. Information-theoretically grounded identifier framework with embedded foreign-key Key Ring, User Tags, and Variant Slots; deployed across StackAdapt, Meta, LinkedIn, TikTok, and Google. Data integrity baked into the encoding layer.
Generative-Engine Optimization (GEO). Designed and shipped agentic-search optimization for the organic marketing pipeline and automated RFP responses, capturing government contract opportunities others were missing.
Direct-Mail Innovation Research. Led technical research and integration design for physical/digital attribution. Evaluated USPS Informed Visibility, PURL systems, BLE beacons, conductive ink, and Digimarc as a portfolio approach to bridging postal infrastructure with downstream digital signal.
Eight-platform abstraction layer. Standardized CS-to-Media handoffs across Meta, Programmatic, YouTube, Search, Native, X, LinkedIn, and Reddit into a single normalized schema and versioned spec, recovering 62 weekly team hours.
Build-vs-buy at scale. Replaced a $50K/yr LiveRamp vendor feature with an Apps Script pipeline that parses 12-13 daily emails, dedupes CSVs, and refreshes every 15 minutes. Same outcome, zero recurring cost.
Marketing Mix Modeling. Google Meridian and Facebook Robyn alongside FFT and Bayesian causal-inference techniques to produce defensible attribution from messy performance data.
Edge-compute AI. CLIP-based image classification and Ollama thin-model deployments on Arduino NICLA and Raspberry Pi, making inference local, cheap, and fast.
Production Claude integration. Arena Regulatus Slack bot deployed via Make.com to the Claude API to Slack, tuned to enforce QA culture and add personality to the media-ops workflow.
Cryptographic campaign-ID hashing. MD5 aggregate key that solved a chronic cross-system join problem (typos and freehand campaign names). Running in production for years and trusted by data, sales, and PM teams alike.
Professional Experience
Director of Engineering & Innovation
Arena Communications (arenawins.com)
- Own AI, engineering, and IT strategy at the executive level: author it, fund it, build it, and lead the whole organization through adopting it. Chair of the internal AI and Automation committees.
- Run the function as a P&L: $1.2M in realized annual value across 25 production systems, valued bottom-up across nine channels with COGS, margin, and EBITDA impact modeled per build; 176 staff hours returned weekly (≈4.4 FTE).
- Lead a tight-knit team of engineers who own their systems end to end; shipped 4 production AI models and the agentic ad-buying agent, every deployment secured, governed, and monitored.
- Selected and rolled out Claude Enterprise and OpenAI Enterprise for 100+ staff, wrote the enterprise AI policy that governs their use, and run the training program behind adoption.
- Automated the workflows that drain an agency: daily compliance sweep across roughly 1,400 ad accounts, four-hour cross-platform reconciliation, client-services-to-media handoffs, and campaign setup pipelines.
- Developed standalone IP alongside operational tooling, including a billable productized service and a proprietary dataset asset, and drove MMM research with Google Meridian, Facebook Robyn, FFT, and Bayesian causal inference.
- Modernized enterprise IT for 100+ employees: quarterly trainings, hardware tokens, IAM hardening across AWS, GCP, and Google Workspace, vendor assessment, and a 24-hour incident protocol.
Executive
AdPlacementPartners
- Built the GEO / agentic-search optimization practice that turned organic marketing into a structured pipeline source and automated RFP responses for government contracts.
- Negotiated and signed agreements with StackAdapt, DeepSync, LiveRamp, Meta, and Google; ongoing engineering liaison influencing vendor roadmaps.
- Stood up the data-lake architecture and self-serve reporting layer; retired $90K/yr in warehousing costs and unlocked downstream monetization.
- Sold $250K+ in technical consulting in a sales-engineer capacity.
Senior Full-Stack Developer
The Professional Education Institute
- Built attention-tracking infrastructure on top of GTM and custom JS to capture mid-webinar engagement and dropoff signals; designed mid-stream interest-meter modals and heat-map-driven retargeting paths.
- Mentored junior developers and modernized sales infrastructure, lifting sales 17% in six months.
- Navigated a 13-framework, 5-language tech-debt landscape without losing delivery cadence.
Consultant (Part-Time, concurrent)
Ad Placement Partners
- Hardened AWS ETL pipelines and Lambda aggregation; invented new digital-ad products and insights driving net-new revenue.
Director of Data & Reporting
Ad Placement Partners
- Led 7 engineers across complex political data sets; coordinated 15 staff across 70 client campaigns at peak.
- Built the proprietary data and reporting system that increased revenue opportunities 120% and ad optimization yield 570%, saving millions in 2022.
- Designed the MD5 campaign-ID hashing layer; built cross-vendor harmonization tables; migrated the on-prem SQL stack to a cloud-native data lake.
- Spun up an email rental and validation system using AWS-based validation, social enrichment, and ML-driven topic propensity, the foundation for a future API-rentable audience product.
Senior Full-Stack Developer
Arena Communications
- Built and maintained sites for U.S. legislators; developed custom .gov data APIs.
- Maintained CI/CD pipelines and pre-election load testing; automated repetitive campaign operations.
- Built a mentoring curriculum that raised junior dev and intern skill quickly.
Freelance Consultant & Developer
Various Clients
- Multi-disciplinary technical resource for e-commerce vendors, hedge funds, and government clients.
- OSINT and threat analysis for government clients; SaaS product strategy and SEO; Shopify migrations and retention strategy.
Tech Stack
Emerging / Frontier. Agentic AI systems, MCP tool design, GEO (generative-engine optimization), MMM (Meridian, Robyn), causal/Bayesian inference, FFT, edge AI (Arduino NICLA, Raspberry Pi, Ollama thin), tokenized AI streaming.
AI Platforms. Claude (Enterprise, API, Bedrock), OpenAI (Enterprise, API), PyTorch, HuggingFace Transformers, CLIP, Ollama, AWS Bedrock, AWS SageMaker, AWS Rekognition.
Cloud. AWS (Lambda, S3, IAM, DynamoDB, Aurora, Athena, API Gateway, EC2, Bedrock, QuickSuite, CloudWatch, LightSail). GCP (Firebase, BigQuery, Kubernetes, Docker, Compute Engine, billing/IAM).
Languages & Web. Python (Torch, Pandas, NumPy, Flask, Selenium), JavaScript (Vanilla, jQuery, Vue, React, Chrome MV3 extensions), Swift (AppKit, SwiftUI, IOKit), PHP (Laravel, Slim), Go, Node.js, ClojureScript, SQL/PostgreSQL/MySQL/SQLite/MongoDB.
Tooling. GitHub, Jira, Monday, Workfront, Jenkins, Make.com, Google Apps Script, VSCode, PowerShell, Bash, Linux (CentOS, Debian), Google Workspace Admin.
Selected Personal Builds
Lidless (open source, MIT)
macOS menu bar app that keeps a MacBook awake with the lid closed and turns itself off the moment you unplug. Swift with no dependencies; IOKit power and lid state, a scoped sudoers rule, a safety watchdog, a CLI, and a universal ad-hoc-signed build from a Makefile.
Elemental Arena, an orb battle game
Deterministic browser battle simulator: nineteen fighter templates, three modes, a roguelite campaign, layered synthesized audio, match recording, and every match reproducible from a seed. Vanilla JS, no build step.
Dual-zone wireless garage indicator
Three ESP32 nodes communicating via ESP-NOW peer-to-peer, HC-SR04 ultrasonic sensors, and a Waveshare 64×64 HUB75E LED matrix. Multi-node embedded with timing-sensitive matrix driving and sensor fusion.
Smart sprinkler controller
Replaced a commercial Orbit unit with a Raspberry Pi and relay board driving 24VAC solenoid valves on OpenSprinkler Pi, integrated with a multi-month lawn-restoration plan.
Home air-quality monitor fleet
ESP32 + SCD41 CO₂ sensor + SSD1306 OLED, with WiFi dashboards, ntfy.sh push alerts, and historical logging to Google Sheets. Full data path from device to dashboard to alert to log.
NIST-style RNG testing kit
Browser-side analogs of frequency, Benford, Monobit, and Runs tests over four generators, with the math shown. Built because "random enough" is not a number.
Education & Certifications
B.S. Information Technology
Brigham Young University, Idaho
Doctoral Candidate, Human-Centered Computing (on hold)
University of Utah. Emphasis: AI Modeling & Quantum Applications
Interests & Research
Contact
For introductions or to discuss opportunities, reach out via LinkedIn or GitHub.
Want to try Pixel Lab or Bid Inspector? DM me on LinkedIn for beta access.