Proof
I don't show client logos. I show my own companies.
These are real businesses I operate AI-first. The playbook I teach is the one I use every day.
PurposeLife
AI Coaching Platform
Before
- ✕Coaching required 1-on-1 human sessions at $150+/hr
- ✕Assessment scoring done manually by trained staff
- ✕Career matching relied on static databases and counselor knowledge
- ✕Scaling meant hiring more coaches linearly
After
- ✓AI handles entire coaching conversations grounded in PhD research
- ✓Automated assessment scoring and career matching
- ✓Platform operates with one founder + AI agents
- ✓Scales to thousands of users without proportional headcount
Stack: Next.js, Anthropic Claude, Prisma, Vercel, Stripe
What still needs humans: Complex life decisions still benefit from human follow-up. AI handles the 80% — structured coaching, scoring, matching. The 20% edge cases get flagged for human review.
Invown
Proptech / RegCF
Before
- ✕Investor outreach: manual research, generic email templates
- ✕Meeting follow-ups: handwritten notes, delayed sends
- ✕Client onboarding: paper forms, back-and-forth emails
- ✕One person could manage ~20 investor relationships
After
- ✓AI-powered research → personalized investor emails → CRM logging
- ✓Fathom → Claude → Gmail pipeline for automated meeting follow-ups
- ✓Automated client onboarding workflows
- ✓One person manages 100+ investor relationships
Stack: Claude, Gmail API, Fathom, CRM integrations
What still needs humans: High-stakes investor negotiations still need a human. AI handles research, drafting, and follow-up — the operational weight. Final review stays with the founder.
Rental Operations
Property Management
Before
- ✕Facebook Marketplace inquiries answered during business hours only
- ✕Tenant screening: manual background checks and reference calls
- ✕Showing scheduling: phone tag across multiple properties
- ✕Evenings and weekends meant missed leads
After
- ✓AI agent handles Marketplace inquiries 24/7
- ✓Automated tenant qualification and pre-screening
- ✓Self-service showing scheduling with automated follow-up
- ✓One person managing multiple properties across 3 CT markets
Stack: Claude, Facebook API, OpenClaw automation
What still needs humans: Lease signing, maintenance emergencies, and complex tenant situations still need a human. AI handles the intake funnel — the highest volume, lowest complexity work.