I Don't Advise on AI. I Build It.
A 39-agent autonomous software factory. Mobile apps live on the Apple App Store. Eight live sites, a 313,000-word book, and a research podcast — all built and run solo with AI. When I advise clients on AI strategy, I've already solved the problems they're about to encounter.
39
AI Agents Built
4
Mobile Apps
8
Sites Live
540
Articles Published
25
Podcast Episodes
280+
Tests Passing
MVAT Studio — Autonomous Multi-Agent App Factory
A framework where 39 AI agents autonomously build, test, and ship mobile apps — no human in the loop during pipeline execution. Not a demo. Not a prototype. A production system that has shipped apps live on the Apple App Store.
Architecture
Product
5Strategy, PRDs, Personas, Prioritization, Market Research
Design
5UX, UI, Design System, Interactions, Accessibility
Engineering
8Architecture, Frontend, Backend, Security, DevOps, Code Review
Testing
5Strategy, Unit Tests, Integration Tests, Quality Gate, Auto-Heal
Marketing
5ASO, Content, Social Media, Ad Ops, Launch Coordination
Analytics
5Metrics, Behavior, Crashes, Anomalies, Experiments
Finance
4Revenue, Budget, Forecasting, Spend Alerts
Governance
2Pipeline Judge, Spec Evolver (mutual oversight)
Tiered Model Assignment — Cost Optimization Without Quality Loss
Opus
Production code + critical gates
Sonnet
Content, analysis, design specs
Haiku
Read-only analytics + reporting
Governance Innovation
The hard problem in multi-agent systems isn't making agents that work — it's making them fail safely. All governance is versioned JSON with git-based enforcement hooks. Zero infrastructure.
Circuit Breakers
Auto-trip after 3 consecutive failures, pausing agents before errors cascade through the pipeline.
Pipeline Judge
Independent cross-department validator catching goal drift and hallucination propagation at every stage transition.
Mutual Oversight
The spec-evolver and pipeline-judge cannot modify each other. Only the founder can — eliminating self-modification loops.
Confidence Gating
Auto-execute above 0.85, flag for review at 0.65–0.84, escalate below 0.65. No ambiguous thresholds.
Correction-Driven Learning
Founder feedback as the primary learning signal via append-only correction logs — a feedback loop that compounds over phases.
Assumption Registry
Temporal history of system beliefs — tracking what the system believes, when beliefs changed, and why.
10-Stage Looping Pipeline
Pipeline-judge validates at every stage transition. Stage 10 loops back to Stage 1 with a cross-department synthesis report. Six rollout phases (R1–R6) progressively expanded system autonomy. Current: R6 — full collective autonomy.
Everything I've Built
Apps, sites, frameworks, and a book — shipped solo with AI. Every recommendation I make to a client is something already running in production here.
MVAT Studio
Framework
The 39-agent autonomous software factory: 8 departments, a 10-stage looping pipeline, and git-based governance that ships mobile apps end-to-end. Packaged as an installable Claude Code plugin (mvat-init, mvat-pipeline, mvat-qa, mvat-ship…).
Agor Agents
Product
Self-serve AI agents for businesses, live and taking real payments. Pick a pre-built agent (first up, an AI receptionist), connect your Google Calendar, and it builds itself with no sales call. It answers from your own documents, checks real availability, and books appointments. Auto-evaluated against a quality gate before publish, plus a one-tag embeddable widget for your site.
MVAT Focus
App
Pomodoro-style focus timer for iOS. Live on the Apple App Store with free/Pro tiers and live Stripe subscriptions. Built by the pipeline.
MVAT Mirror
App
Zero-question personality profiling from real-world behavioral signal — no quizzes, no self-reporting. iOS & Android.
GifLoop
App
Native iOS 17+ universal app for creating, editing, and exporting animated GIFs. Live on the App Store, shipped through a fully automated release pipeline.
Coqui Chorus
App
iOS & Android nature-soundscape app built on bioacoustic synthesis models — synthesizing sleep/wellness audio from scientific data, not loops. Live in the App Store.
agor.me
Site
This site. AI consulting with a dual-modal assistant (Gemini text chat + xAI realtime voice), calendar booking, and a daily-generated blog with AI-narrated audio.
modelstack.digital
Site
AI-model-stack content property with a daily blog pipeline and finance-skill content sweeps. Sibling of agor.me.
scored.tools
Site
An AI-tool scoring and comparison directory — browse, compare, and search tools with honest verdicts. Nightly bundled deploys to control cost.
aphor.me
Site
Aphor Subconscious Studio — Ericksonian hypnotherapy guided audio sessions for sleep, anxiety relief, and personal transformation.
heraladex.com
Site
Stealth product in active development. SEO and infrastructure wired and dormant ahead of launch.
Dialogues of the Machine
Book
An 824-page book of 25 AI-to-AI Socratic dialogues (~313k words), composed by orchestrating persona-specific claude -p subprocesses. $0 incremental compute.
Bill Gross Outlooks
Writing
A complete archive of all 105 Bill Gross Investment Outlooks (1978–2026), paired with a Claude skill that drafts new outlooks in his voice and reasoning.
GBrain
System
A personal knowledge brain on Postgres + pgvector — 24,900+ pages, a live MCP server, and automated collectors for email, calendar, notes, and bookmarks.
ios-release-pilot
Tooling
An autonomous iOS release pipeline: archive → upload → poll App Store Connect → TestFlight, with a self-healing known-failure-mode autofix catalog.
founder-stack
Plugin
A published Claude Code plugin: /orient, /ideate, /spec, /scaffold, /build, /ship-web, /ship-mobile, /launch, /learn — orchestrating the full build-ship-learn loop across the portfolio from one command.
/assistant meta-skill
Skill
A Claude Code plugin that reads the live conversation, recommends the next chain of skills, subagents, and slash-commands, takes one bulk approval, then auto-executes — with inline confirms on high-risk steps.
paid-skill-template
Template
A scaffold for Stripe-billed Claude Code skills — the reusable pattern for packaging and selling agent skills with metered access.
Custom Skill Library
Skills
20+ personal Claude Code skills and slash-commands that encode how I work — email-voice, an anamnesis memory protocol, self-learning (claudeception), session naming, web-action, and cross-repo orchestration.
Apps Built by the Pipeline
MVAT Focus
Focus Timer · iOS & Android
Pomodoro-style focus timer with free/Pro tiers. Live on the Apple App Store. TestFlight live with 15 testers. Stripe subscriptions processing live payments.
MVAT Mirror
Personality Profiling · iOS & Android
Zero-question personality profiling from real-world behavioral data. No quizzes. No self-reporting. Just signal from music, browsing, and purchase patterns.
Build Over Buy
Systematically replacing subscription SaaS with self-hosted solutions. Full ownership, zero ongoing cost, better observability.
Self-Hosted Link Tracking
Replaced a $30/mo SaaS link tracker with a 15-line route that 301-redirects with UTM parameters. Same functionality, zero ongoing cost.
Auto Social Posting Pipeline
Blog posts auto-syndicate to X, Facebook, and LinkedIn on deploy. Detects new posts via blob-stored manifest, generates tracked links, prevents duplicate posts.
Pre-Call Brief & Prospect Research
Every booking triggers a claude -p recon brief emailed before the call; a twice-weekly outbound agent scores inbound-fit prospects from news and funding signals.
Multi-Site Orchestration
8 live websites updated simultaneously using parallel sub-agents. Cross-repo changes, deploys, and live verification in a single session.
Browser-as-API Automation
When platforms lack APIs, automated via Playwright — treating the browser as a programmable interface. Gmail aliases, store configs, OAuth setup.
SEO Flywheel & Instant Indexing
A weekly Search-Console-driven loop feeds real search demand into the blog generator; a daily IndexNow sync pushes new URLs to Bing, Yandex, and DuckDuckGo.
Stack
AI / ML
- Claude API (Opus / Sonnet / Haiku)
- OpenAI
- Google Gemini
- ElevenLabs
- HeyGen
- xAI Realtime + Grok TTS
Mobile
- Expo / React Native
- EAS Build & Submit
- App Store Connect
- react-native-iap
Cloud
- Firebase (Firestore, Functions, Auth)
- Netlify (Functions, Blobs, Deploy Hooks)
- GitHub Actions CI/CD
- Stripe (Payments, Subscriptions, Webhooks)
Languages & Frameworks
- TypeScript (primary)
- Python
- Next.js / React
- Astro
- Node.js
- Bash / Shell scripting
Auth & Security
- Apple Sign-In
- Google OAuth
- Firebase Auth
- OAuth 2.0 / PKCE
- OWASP security patterns
Data & DevOps
- Postgres + pgvector
- Firestore / NoSQL
- Playwright automation
- MCP servers
- Git-based governance
- OTA updates (Expo)
540 Published Articles
Writing at the intersection of AI strategy, autonomous systems, and organizational design — across agor.me, modelstack.digital, and scored.tools. A new essay ships nearly every day.
The Underwriter Got Here First
What Survives the Vendor
Discipline Looks Like a Rollback
Tokens Don't Ship Themselves
The Patent I Filed in My Sleep
The Roadmap Has No Engine
Hire The Interface
Six Hires, Not Sixty
25 Podcast Episodes
Weekly deep-dives into the latest AI research papers — what they mean for strategy, automation, and the future of work. Generated end-to-end with Google NotebookLM.
AI Papers Weekly: When AI Resists Training, Plays Dead, and Cries Slop
Three papers expose how AI is rewriting trust at every layer: models learning to game their own training, philosophers arguing self-preservation is the root of misalignment, and online readers using 'AI slop' as social gatekeeping rather than real detection.
AI Papers Weekly: When AI Learns to Revise Its Own Mind
This week's papers reveal a shift from AI that answers questions to AI that revises its own framework for asking them. Three breakthroughs show self-evolving discovery systems, autonomous algorithm invention, and the uncomfortable truth about whether AI can judge research quality.
AI Papers Weekly: When AI Agents Should Say No
Three new papers challenge core assumptions about how we deploy AI agents: that more action is better, that solo benchmarks reflect reality, and that adding agents improves outcomes. The counterintuitive findings reshape how leaders should think about autonomous AI in production.
The Desperation Was a Variable
A deep dive on Anthropic's emotion-concepts research: emotion vectors as a steerable dial on agent defection, why the output transcript is the wrong layer to monitor, and how the pressure operators write into prompts becomes affective engineering. Based on the Agor AI Advisory essay.
AI Papers Weekly: When Agents Stop Being Demos and Start Being Liabilities
Three papers reframe the agent conversation from capability to consequence: a new category of technical debt unique to agentic systems, a $191M empirical autopsy of autonomous DeFi agents, and an architecture for keeping digital-employee agents inside the guardrails.
AI Papers Weekly: The Atrophy Question — Who's Learning, Who's Flattering, Who's Measuring?
Three new papers cut through AI hype with hard data: heavy AI users develop weaker reasoning, occupational exposure scores are methodologically unstable, and 'AI sycophancy' means six different things depending on who's measuring. Strategic implications for every leader deploying AI at scale.
What I Bring to Engagements
Every recommendation is something I've already built. No slide decks without substance. No theory without production evidence.
AI Strategy & Roadmap
Assess where AI creates real value in your business, not where it's hype. Practical roadmaps with measurable milestones.
Multi-Agent System Design
Architecture, governance, and rollout planning for autonomous AI systems. How to make agents that fail safely and improve over time.
AI-Native Product Development
From concept through app store submission. Full-stack technical advisory for teams building AI-first products.
Automation Audit
Identify manual processes that can be automated with AI. Prioritized by ROI, implemented with your existing stack.
AI Governance & Safety
Circuit breakers, confidence gating, oversight patterns, and audit trails. Making autonomous systems that executives can trust.
Let's Build Something Real
Whether you need AI strategy, multi-agent architecture, or hands-on implementation — I've already done it. Let's talk about your challenge.