A behavioral intelligence platform for couples I built end to end and operate in production. Three-stage RAG pipeline over 8,000+ embedded research chunks, 33 behavioral analysis capabilities, voice AI, and production eval gates. Product, backend, payments, iOS pipeline, and deploys: all mine.
Claude + OpenAI · RAG / vector embeddings · Supabase / Postgres / RLS · Stripe · iOS + web · Cloudflare
A fleet of autonomous AI agents running 24/7 on VPS infrastructure for my own businesses. Research, content drafting, scheduled jobs, and a chat command center, with multi-model orchestration across providers.
Multi-agent orchestration · VPS / Docker · Cron + webhooks · Telegram ops
An AI support and coaching agent answering real customers inside a client's education platform. I built it, deployed it, and operate it like an SRE: root-causing outages, fixing timeout chains, and resolving tickets when it breaks. Building agents is easy. Running them in production is the skill.
n8n workflows · Claude · Cloudflare Pages · Supabase · Incident response
A two-agent system for a real estate appraisal firm: one agent scrapes county, MLS, and tax sources in parallel; a second QC agent reviews every field against source data. Cut the owner's review time by roughly half and opened a clear path to doubling monthly capacity.
Parallel scraping agents · QC / validation agent · Structured output
For a senior performance executive at a world-class sports organization: the AI transformation of a comprehensive high performance audit instrument, from a manual workbook into a scalable platform design with benchmarking and trend analysis.
AI product architecture · Assessment analytics · Executive advisory
A complete AI spokesperson system for a coach: cloned voice, an approved stylized avatar built from real photos, and lip-synced video generation. Five-minute videos at a small fraction of filming cost, with a proven per-video cost model.
Voice cloning · Image + video generation · Lip sync · Cost engineering
Every AI initiative dies the same way: the tools work and the people don't change. Adoption is a behavior problem. That is the problem I trained for.
Executive and team sessions on what agentic AI actually means in practice. Not prompts. Systems that act on your behalf, and what it takes for a team to trust them.
Certified in high performance coaching and relational intelligence. Years of coaching operators and couples through the hardest behavior changes there are.
Onboarding flows, retention loops, and habit design inside products I run. I don't theorize about adoption curves. I watch mine every week and tune them.
I spent 8 years in Wall Street cybersecurity, working up from help desk to security engineer at a law firm. Threat detection, incident response, compliance. My job was protecting systems from external threats.
I left to build my own businesses and built a $10M real estate portfolio. Then I got certified in high performance coaching and spent years coaching entrepreneurs and couples. That work taught me the other threat model: the internal patterns that make smart people resist the very change they asked for.
When agentic AI arrived, my two careers turned out to be one career. Threat detection became pattern detection. Incident response became intervention design. Compliance became daily practice. Same architecture, different domain.
So I built. A live App Store product with a full RAG pipeline. Agent fleets that run my businesses while I sleep. Client systems I operate in production, including the 2am debugging. And workshops that turn skeptical teams into adopters.
That combination is the job description of AI transformation. Most builders can't drive adoption. Most adoption people can't build. I do both, and I have the shipped systems to prove it.
Use-case discovery, build-vs-buy calls, and a sequenced roadmap tied to business outcomes, not demos.
Production agent design, retrieval pipelines, vector search, evals, and the guardrails that keep them honest.
Mapping real processes into automations people keep using. n8n, APIs, and the glue between systems.
Executive sessions and hands-on team training that move people from anxious to fluent.
Behavior-change design for rollouts: champions, habit loops, and the metrics that show real usage.
Eight years of enterprise security instincts applied to AI: data boundaries, access, risk, and compliance posture.
Whether you need a fractional AI lead, a transformation partner, or a builder who can also win over your team, the first step is a conversation.
