A builder and security researcher who came up through the creator economy.

How I work

The through-line is ownership. Hand me a problem and the room to own it, and I architect it, direct the build with AI agents, ship it, and stand behind it. That is how the live projects got built and how the open-source fixes get written.

I am not a single-function specialist. I do my best work where one person owns the whole thing: the product call, the code that ships it, and the security review that keeps it standing.

I am open to contract and project work, remote or Northern California. For paid security work, secure code review and vulnerability fixes run through Fiverr.

Story

I started at Twitch in 2016 on the customer-facing side. CRM and customer service tickets during a stretch when the platform was growing faster than its support team. It sounds like the wrong starting point for an influencer marketer, but it gave me a ground-level read on creators and viewers that most marketing leads have to learn from a deck.

By 2019 I'd moved into influencer marketing and stayed there for the next six years. I ran tentpole creator programs for the NFL, the Olympics, and Amazon Prime Video. Multi-region, multi-brand, real budgets, real KPIs. Talent negotiation, content frameworks, paid amplification, performance reporting. The whole stack.

The part of the job I'm proudest of, and the part nobody outside the org knew about, was the tooling. I built AI systems inside Twitch's marketing team before generative AI was on most marketing leaders' radars. A creator pricing chatbot the legal and finance teams ended up using. An automated contract generator that cut errors and turnaround across cross-functional teams. They weren't side projects. They were the infrastructure my team ran on.

I left Amazon in October 2025. The bet was simple: with the AI tooling that existed at that moment, one operator could deliver senior marketing work at a level that used to require a team of five, and could ship their own software while doing it. Six months in, the bet is paying.

Willhite Strategy Group is the consultancy and the platform underneath it. The platform runs a 467K-creator CRM with multi-stage AI scoring, automated outreach, and a client portal. Six integrated subsystems that share data and feed each other. CommQuest, the content agency I started in 2021, is the second income line and proof that the model works across content production too. Brands include Guinness World Records, Vasconia, and PROOF Wallets.

On the side, I ship products. Dealophant is a price-per-unit Amazon shopping engine with a B2B brand marketplace bolted onto it, live since February. My GLP Shot is a privacy-first GLP-1 injection tracker with end-to-end encrypted cloud sync. I have a Chrome extension, an OSINT-y creator scoring engine, a Telegram-bot ops EA, a self-hosted Reddit agent, and a decision-learning service that trains itself on every creator I vet. Forty plus shipped projects.

I direct AI agents to build and secure production software, the way most senior PMs learned to direct engineering teams. That advantage compounds every quarter.

Roles

40+ shipped projects in six months

Since leaving Amazon I've directed the build of an integrated software platform spanning three VPS servers, two businesses, and a portfolio of consumer products, all without writing code from scratch. I architect, decide, and ship; AI agents do the typing.

Flagship products

Tools I've shipped to a self-hosted internal portal

All gated behind one auth at tools.willhitestrategy.org:

• Sports Tracker: UFC + USL League One schedule. Daily Telegram alerts. Web-push (VAPID) browser notifications. PWA for iOS Safari support.
• Audio Transcriber: drop-in audio to transcript + AI synopsis via Gemini 2.5 Flash.
• Peptide Calculator: reconstitution math, dosing schedules, injection planner.
• EXIF Remover, Image Resizer, Video Downloader (self-hosted Cobalt), File Transfer.
• Tracking Links System: wsgx.cc short-link domain with two link types (campaign vs. resource), priority routing, NocoDB CRUD.
• Trading Bot, dual-engine sandbox: Freqtrade for crypto + custom Python engine for stocks (Alpaca paper trading, MomentumMeanReversion strategy with ADX regime detection).

AI & agent systems

• Sam & Hermes: Telegram-bot ops EA + notification dispatcher across the entire infra.
• Professor: decision-learning service that mirrors every creator vetting approve / decline / skip and trains a baseline ranker, transitioning to LightGBM at ~500 decisions.
• CommQuest Inbox Agent: hourly Gmail monitor that auto-classifies brand pitches and drafts replies.
• Sentinel: daily SEO + brand-mention tracking across ChatGPT, Claude, Perplexity, and Gemini for client brands.
• WSGiQ Chrome Extension v1.2: iframe-bridge extension that overlays the WSG vetting UI onto creator pages on TikTok, IG, and YouTube, with a multi-creator scan grid.
• Marketing Intelligence Hub, self-hosted okara.ai equivalent: AI Humanizer, Reddit Agent, GEO Score Tracker, SEO Agent.

Infrastructure I run

Three VPS servers, all self-managed. VPS1 is the workhorse (admin, NocoDB, Dealophant, My GLP Shot, Tools Portal, Sports Tracker, Trading Bot). VPS2 hosts the WSG back-office (Refinery, Prowler, Camofox stealth-browser, SearXNG, Karakeep, Meridian, CommQuest Inbox Agent, Mattermost, Planka, Keila, OpnForm, Uptime Kuma). VPS3 is dedicated to client websites on CapRover + Docker Swarm.

I built wsg-infra, a CLI that auto-collects ports, services, and resources from all three VPS every 10 minutes and aggregates to a single JSON, plus a parallel wsg-infra-audit git repo that tracks historical state with a CHANGELOG, so any infra change is forensically reconstructable from git log.

Self-hosted OSS replacing recurring SaaS bills

Slack → Mattermost. Trello → Planka. Mailchimp → Keila. Todoist → Super Productivity. Coolify → CapRover. Google Analytics → Umami. okara.ai → custom Marketing Intelligence Hub. Every recurring bill became a hard look at OSS alternatives.

Stack

Operating principles

• Architect integrated systems, not isolated apps. Six WSG subsystems share data, share AI services, and feed each other. A Marketing Hub Reddit hit becomes a Sales Hub deal. A creator declined in Client Portal teaches the Creator Engine. The compounding is the product.
• Cost discipline through architecture. All AI calls route through one cost-tracking helper. Two-tier (regex → AI fallback) parsers to keep token spend low. Free-tier OpenRouter model chains with paid fallbacks. Self-host before subscribing.
• Forensic infrastructure hygiene. Live state (wsg-infra) and historical state (wsg-infra-audit) are both queryable. Every infra change updates a CHANGELOG in the same commit.
• Direction over implementation. I make architecture calls, set constraints, and run agents. I'm responsible for what gets built and why. The code gets typed by AI. I'm the operator.
• Ship-first, refactor when it hurts. Working systems beat planned ones. I'd rather have the v1 in production teaching me what the v2 should be.

Let's talk

Best way to reach me is email or LinkedIn. I'm fastest on email. If you have a project, a contract engagement in mind, or want to talk through whether what I've built could do something inside your org, let's set up time.