Products
Things I built because they
needed to exist.
These aren't consulting projects. They're things I built because I saw a problem worth solving and decided to solve it.
Platform
Tenet
Agent-native local discovery.
The way people discover local businesses is changing. AI assistants are becoming the first point of contact — and they need structured, trustworthy data to work with.
Today's AI tools are mostly guessing. They hallucinate business names, invent menu items, and pull from outdated aggregators. There's no structured, agent-queryable source of truth for local business data.
Tenet is infrastructure for agent-native local discovery — purpose-built so AI assistants can find, understand, and recommend local businesses with real data behind every answer. It speaks the language of AI agents through open protocols like MCP (Model Context Protocol) and A2A (Agent-to-Agent), so any AI assistant — Claude, ChatGPT, custom copilots — can query directory data in real time.
The platform is designed to power independent discovery directories across multiple verticals. Each launches as its own brand, built on shared infrastructure that's already proven in production.
Agent Protocols
MCP and A2A support so any AI assistant can query directory data natively — no scraping, no hallucination.
Multi-Vertical
Shared infrastructure powers independent directories across dining, professional services, healthcare, legal, and home services.
Structured Data
Curated, verified business data — menus, pricing, hours, services — organized so AI agents can use it directly.
Real Answers
When an agent queries Tenet, it gets verified facts — not guesses. That's the difference between an AI that knows and one that hallucinates.
Built on Tenet
Misen.ai
Everything in its place.
A curated dining directory for the Minneapolis-St. Paul metro — built so AI agents can recommend restaurants with real data, not guesses.
The name comes from mise en place — everything in its place — and that's exactly what Misen does for restaurant data. Structured menus, dietary details, pricing, reservation availability, and neighborhood context, all organized so AI agents can use it natively.
When an agent asks "What's on the menu at Bar La Grassa?", it gets back structured data: every dish, every price, dietary flags, reservation platform, and more. That's the difference between an AI that guesses and one that knows.
Misen is live today and registering with major AI tool directories. It's the first vertical on the Tenet platform — proof that agent-native discovery works in practice, not just in theory.
How it works
Curated data, not scraped data
Restaurant information is verified and structured — menus, prices, dietary options, hours, reservation links — not pulled from stale aggregators.
Agent-native protocols
AI assistants query Misen via MCP and A2A — the same protocols they use for other tools. No special integration needed.
Real answers, not hallucinations
Every recommendation is backed by verified data. The agent knows what's actually on the menu, what it costs, and whether you can get a table.
misen.ai — live now, covering Minneapolis-St. Paul dining. Read the case study →
Product
Surfaced
Be where they're looking.
Ongoing SEO is a continuous job. Most small businesses can't do it. This platform does it for them — automatically.
The system monitors rankings, analyzes competitors, generates content, publishes to Google Business Profile, responds to reviews, audits technical health, and measures what's working — all without the business owner doing anything beyond approving what it produces.
It started as a project for one client. It's now a multi-tenant platform running across multiple businesses with 10 active automated pipelines. Read the full story →
Data Intelligence
Twice-weekly rank tracking, weekly Search Console and GA4 analysis, technical health audits, and monthly competitor keyword discovery.
AI Content
GBP posts 3x/week, real-time review responses with sentiment-aware tone, and long-form content generated from keyword gap analysis.
Human-in-the-Loop
One-click email approvals with no login required. Configurable thresholds per client. Full audit trail of every action and outcome.
Closed-Loop Measurement
Every action is tracked back to its outcome. The system measures what actually moved the needle — and learns from it.
How it works
Data flows in automatically
Rankings, search performance, traffic, technical health, and competitor data — collected on schedule without any manual triggers.
AI generates and recommends
Content is drafted, opportunities are scored, and actions are proposed — each with the reasoning behind it.
You approve with one tap
An email arrives with a preview. Approve or reject. No login, no dashboard, no app.
The system measures and learns
Every action is tracked to its outcome. What worked gets reinforced. What didn't gets adjusted.
The platform is built on a multi-tenant architecture — adding a new client is configuration, not code. The template system is designed to extend beyond SEO into other verticals.
surfaced.sh — operational now. Read the case study →
Product
Airloom
AI Voice Biographer
Everyone has stories. Most of them disappear.
Not because people don't want to preserve them — but because sitting down to write a memoir is a project, and projects require time, structure, and sustained motivation that most people never quite find. The stories stay in people's heads until they don't.
Airloom is a voice-first AI biographer that pulls your life stories out through conversation.
You don't write anything. You talk. Airloom interviews you — asking questions, following threads, drawing out details you might not have thought to include on your own. It knows when to push deeper and when to move on. The conversation feels natural because it's designed to.
Home
Conversations
Time Capsule
Audio Storyteller
Your stories narrated and produced into a finished audio experience.
Video Montage
Your narrative paired with photos and visuals.
Published Book
A complete memoir, structured and ready to print. (In development)
The people I had in mind when building this: adult children who want to capture their parents' stories before they can't. People who've lived unusual lives and always meant to write them down. Anyone who has ever thought "I should record this before it's gone" and never found the time.
Airloom finds the time for you.
Product
Rapport
Voice-First Micro CRM
Most CRM tools are built for sales teams with dedicated time for data entry. They assume you'll sit down after a meeting, open the app, fill in the fields, log the notes, set the follow-ups.
Most people don't do that. The meeting ends, something else starts, and by the end of the day the details are fuzzy. The follow-up that should have happened Tuesday happens Friday, or not at all.
Rapport is built around a different assumption: the best time to capture a client interaction is immediately after it happens, and the fastest way to do that is to talk.
Home
Actions
Profile
How it works
Speak a 30-second voice note
What was discussed, what was decided, what needs to happen next.
Rapport does everything else
Transcribes, extracts key info, logs the interaction against the right contact.
Follow-ups are queued automatically
By the time your phone is back in your pocket, the CRM is updated and next steps are set.
No typing. No navigating menus. No convincing yourself you'll do it later.
Rapport is designed for the way client relationships actually work — in hallways, parking lots, and two minutes before the next thing starts.
Interested in early access?
These products are in active development. If any of them sound like something you need, I'd love to hear from you.
Get in touch