featured product
Vora
Implemented
An AI CRM platform for service businesses: missed-call text-back, lead follow-up, campaigns, and scheduling in one system. Approval requirements depend on the action and the configured policy.
Your problem. My system.
Every row comes from a real business. Open one to see the fix I build for it.
You're with a client, three calls go unanswered. One was a $3K lead who booked with a competitor by lunch.
AI picks up every call
Voice agent answers instantly, books the appointment, routes emergencies, handles FAQs. 24/7.
Website forms pile up in your inbox. That hot lead from 8 AM? Still waiting at noon. They've moved on.
60-second follow-up
AI scores the lead, sends a personalized response, adds them to a nurture sequence. Automatically.
An hour a day on invoices, calendar updates, data entry. It's the work you hate most and it never ends.
Admin runs itself
Invoices auto-generate from completed work. Calendar syncs from the booking agent. Data entry eliminated.
You know you should post, send that newsletter, update the website. But you're too exhausted by 5 PM.
Campaigns run while you sleep
AI posts to social, sends email/SMS campaigns, writes blog posts, handles SEO. You review over coffee.
You need a receptionist, a marketing person, someone to chase invoices. But payroll for three people isn't realistic.
Coverage without a new hire
Virtual staff for front desk, social media, collections, customer service. They work 24/7 and never call in sick.
Urgent requests buried under spam. Client questions unanswered. You spend 45 minutes just sorting.
Inbox triaged overnight
AI categorizes, auto-replies to common questions, flags what actually needs you. 3 items, not 47.
Intake forms, contracts, and PDFs sit in a queue. Someone has to read each one, pull the fields, and file it. That someone is expensive and slow.
Documents processed on arrival
An agent reads each document, extracts the fields, and files it to the right record. A person reviews the exceptions, not the whole stack.
Every new client means the same emails, the same forms, the same chasing for missing information. It stalls the work and it looks unpolished.
Onboarding that runs itself
The system collects intake, checks it for gaps, chases what is missing, and sets up the account, so the first real conversation is about the work.
A client asks a question and the answer is in a file from two years ago that nobody can find. So it gets re-answered from scratch, sometimes wrong.
Answers from your own records
A knowledge assistant searches your documents and answers with the source attached, so the team stops re-deriving what the firm already knew.
Map the work first. Then automate it.
Most AI projects fail because they automate a process nobody understood. Mine start with lean process mapping, so the AI lands where it pays. I call the method Charted: map the work, gate the risky parts, and keep checking the output after it ships.
Discover
1 wkShort interviews with you and the people who actually do the work. Where does the time really go?
Map
1 wkOne working session. Your core process goes up on the wall: every step, handoff, and workaround.
Prioritize
3-4 daysEvery automation idea gets scored on payoff, feasibility, data, and risk. Only the honest ones survive.
Pilot
4-8 wksI build the winner with one success metric and guardrails attached: anything risky waits for your approval before it runs. We agree up front what scale, fix, or stop looks like.
Scale
1-2 wksYour team gets trained, the runbook gets written, and the system becomes yours to keep.
Assure
monthlyI keep checking the work. Every month I sample what the AI produced, score it against the standard we set, and send a one-page report. Drift gets caught before you feel it.
Phases 1 to 3 are the audit. Phase 4 is the build sprint. Phase 5 closes either one. Phase 6 is the retainer, and it's optional: the system is yours either way.
What I actually build
Four kinds of systems, built from the same working parts.
voice
Voice agents
Answer every call, book the job, route the emergency.
knowledge
Knowledge assistants
Answer from your own records instead of a generic guess.
workflow
Workflow automation
Invoices, scheduling, and data entry that run themselves.
campaigns
Campaign systems
Email and SMS that go out on schedule while you work.
Three ways to work with me
Fixed scope, plain deliverables, and nothing gets built before the audit says it's worth building. Every engagement is a fixed fee, quoted before we start. No hourly billing, no surprise scope. If the scope changes once we're underway, that gets re-quoted and agreed before work continues.
Start here
Process & AI audit
Two to three weeks inside your business. I interview your team, map how the work really moves, and score where AI genuinely pays off (and where it doesn't).
- Your core process, mapped and annotated for automation
- A scored shortlist of AI opportunities, ranked by payoff and risk
- A 90-day roadmap and one recommended pilot
Audits start at $750, fixed scope. You have the number before the call, and build pricing comes from what the map finds.
2-3 wksLet's talk
Build sprint
The top item on your roadmap, built and wired into your real systems. An automation, a website that converts, search visibility, or follow-up marketing. One success metric, your team trained on it.
4-8 wksAssurance retainer
I stay on to check the work. Every month I sample what the AI produced, score it against the standard we set in the pilot, and send you a one-page report. Drift gets caught before you feel it, and you still get tuning and one new automation a month. Capacity is capped at a few clients at a time.
monthlywhat the audit produces
The three templates your audit fills in. You keep all of them, whether or not I build anything.
The sprint and the retainer build on what the audit finds. That order is the point.
The systems I recommend are ones I build and run
I'm a builder first. This work ranges from products I run day to day to prototypes and developer previews whose limits are stated beside the example.
developer tools
Prism
Developer preview
A developer preview for bounded local agent execution, with an inspectable event trace, a contract-derived terminal run record, and fixed limits clearly separated from measured usage.
Agent Team
Implemented
A software agent workflow for planning, implementation, and review. Its example separates a successful run, a retry, and a blocked result, with the default automatic merge policy stated beside the steps.
Agent Config
Implemented
A shared configuration workflow that turns authored fragments into instructions for different coding tools. The example compares generated outputs and shows what becomes stale when a source fragment changes.
Control Center
Prototype
An operations dashboard combining unattended runs, pending decisions, sample spend, and deployment records. This prototype example shows how an operator can find work that needs attention without implying verified deployment.
more products and tools
ChapterHQ
live
A management platform for clubs and nonprofits: members, dues, events, and an AI assistant that answers from the org’s own records.
Site assistant
live
The chat agent running live in the corner of this site, answering questions from my own content.
Open source
live
Tooling I publish on GitHub.
Real Estate Maite
in development
An AI operating system for real estate agents: a team of agents handling follow-up, listings, and paperwork over web and SMS.
Explore the work
The chat assistant on this site is one of these systems. Open it and ask what AI could take off your plate.
On the board this week
Running Vora, a CRM for service businesses
built for
Small businesses of every kind. Service businesses that run on calls, crews, and schedules. Office and professional firms that run on clients, documents, and billing. Any business where the owner does work a system should.
What could AI handle for you?
Tell the assistant about your business and get a plain-language breakdown of what's possible. It's a system I built, working right there in the corner.
Rather talk to me directly? Let's talk · or take the free AI readiness assessment
start here
the method: map it, score it, then automate what pays