You don't need another chatbot. You need a system that answers the phone at 9pm, books the job, audits the vendor invoice, and updates your CRM before you've parked the truck. I build live voice agents, automated document pipelines, and outbound workflows that give you back 20+ hours a week — for operators anywhere, run from Honolulu.
Back-office time recovered per week
Missed inbound leads after hours
From audit to system in production
Live Demo
Two agents currently in production. Real calls, real bookings — not a scripted sizzle reel. Pick a line and press play.
Property Line · After-hours showing request
Inbound buyer · 0:48 · Vapi + Twilio
Transcript snippet: “Aloha, thanks for calling — I can get you on the calendar for that Kaka'ako listing. Are you thinking this weekend or a weekday evening?”
Average response time
1.2s
Sub-1.5 second first-word latency. Callers don't hear dead air, so they don't hang up on it.
Booking calendar sync
Real-time
Writes straight into Google Calendar, Calendly, or your POS. Confirmation SMS fires before the call ends — no double-bookings.
After-hours call capture
98%
Nights, weekends, mid-rush. Every call answered, qualified, logged, and routed — the 2% are hang-ups before ring one.
Want to stress-test it yourself? Book the audit and I'll point a live agent at your actual use case — you call in and try to break it.
Get the demo lineCore Systems
Built on the same stack I run in my own businesses — Vapi, Twilio, Make.com, and Claude. Scoped in the audit, shipped in weeks, owned by you.
A phone line that never sends a paying customer to voicemail. Answers in your brand's voice, qualifies the lead, and closes the booking while you're on the job.
Vapi · Twilio · ElevenLabs
Stop paying someone to retype PDFs. Invoices, contracts, and vendor sheets get read, checked against history, and pushed into your books automatically.
Claude · OCR · Google Sheets · QuickBooks
The connective tissue between your tools. Multi-step pipelines that move data, chase leads, and follow up without anyone remembering to do it.
Make.com · Airtable · HubSpot · Slack
Proof of Concept
My own businesses were the first client. A food truck losing catering leads during the lunch rush, and a small property portfolio drowning in vendor paperwork. Here's the before and after.
Rush-hour calls went to voicemail
Both hands on the flat-top from 11 to 2. Every catering inquiry that called during peak hours hit voicemail — and most never called back.
Vendor invoices typed in by hand
Sunday nights keying distributor invoices into a spreadsheet. Price increases got noticed a month late, if at all.
Leads lived in three places
Texts, DMs, and a notebook on the dash. Follow-up depended on whoever remembered.
Property paperwork by inbox search
Lease dates, renewals, and maintenance approvals tracked by scrolling email at midnight.
Zero missed leads
Voice agent answers every call in ~1.2s, quotes standard catering packages, checks the calendar, and books the date. Deposit link texted before hang-up.
Instant distributor audits
Invoices forwarded to an inbox get parsed line-by-line, compared against the last 90 days, and flagged the same morning a price moves.
One pipeline, one source of truth
Calls, forms, and DMs land in the same CRM with the transcript attached. Follow-up sequences fire on their own.
Property docs that surface themselves
Leases parsed on upload. Renewal and inspection dates pushed to calendar with a reminder chain — no inbox archaeology.
“The stack paid for itself the first weekend it caught a catering booking I would have slept through.”
About
I don't sell theoretical AI. I build the exact systems that run my own businesses every day.
I came up through military financial technology, where the systems I worked on had to reconcile real money under audit — no "move fast and break things" when the numbers belong to someone else. I carried that into an engineering career building production software, and then into my own companies: a food truck and a small property portfolio. I know what it costs when the phone rings during a rush and nobody picks up. I know what a Sunday night of manual invoice entry feels like. That's the gap most AI vendors never see — they've never had to make payroll off a workflow.
So when I scope a system, I start where you actually lose money: the missed call, the untracked lead, the vendor overcharge nobody caught. Then I build the shortest path to killing that problem — usually a voice agent, a document pipeline, and a Make.com scenario doing the boring work at 3am.
I'm based in Honolulu and work with operators anywhere — most of what I build runs remotely by design, and I've never needed to be in the room to ship it. Small businesses everywhere run on the same three things: relationships, thin margins, and a labor market where you can't just hire your way out of an ops problem. The systems I build respect that. You own the accounts, the automations, and the data. No black boxes, no lock-in, no retainer for the privilege of logging in.
Military financial tech
Financial systems and reporting where accuracy is audited, not assumed
Engineering background
Production software, LLM integrations, and data pipelines
Entrepreneurship
Owner-operator of a food truck and a property portfolio — payroll, vendors, and all
The stack I build on
Everything runs in accounts you own. If we part ways, the systems keep working.
Automation Audit
Tell me where your week disappears. I'll tell you whether AI actually fixes it, what it would take to build, and roughly what it costs. If automation isn't the answer, I'll say that too.
Map the leak
We walk your current process and find where hours and leads are actually going.
Scope the system
You get a one-page build plan: what gets automated, in what order, with a fixed price.
Ship in ~14 days
First system live in two weeks, in your accounts, with documentation and a handoff walkthrough.
Four fields. I reply personally within one business day.
I'll reach out within one business day with a couple of times. In the meantime, think about the one task you'd hand off first.
