AI agents for small local businesses
Your business,
on autopilot.
We build AI agents that answer enquiries, fill your calendar, send the quotes, and keep the paperwork moving — so the repetitive work stops eating your week.
ncode — agent session
$ ncode run booking-agent
→ reading inbox & DMs… 14 booking requests
→ checking the calendar… done
→ confirming & sending reminders… done
→ flagging 1 conflict for you
✓ 13 bookings confirmed · 2h 10m saved
$
What we build
// capabilities/01Autonomous agentsGoal in, finished work out. Agents that plan the steps, do the work inside the tools you already use, and check their own output before anything ships.
/02Workflow automationBookings, reminders, quotes, listings, follow-ups — the recurring work that eats your evenings, handled on a schedule or a trigger. You approve only what matters.
/03Deep integrationsWired into your email, calendar, messengers, and payment tools — agents act inside what you already use. No new dashboard to check.
Watch an agent work
// live demoEvery system we ship runs the same loop: understand the task, act with tools, verify its own work. Here it is — slowed down enough to watch.
src: inbox & DMs
BK-0412
read_dm
calendar
send_sms
{}AGENT
dst: calendar booked
BK-0412✓
BK-0412✓
BK-0412✓
⚑ you approve
conflicts go to you
BK-0412
BK-0412✓
BK-0412
BK-0412✓
BK-0412
BK-0412✓
BK-0412
BK-0412⚑
$ booking-agent: reads requests → checks the calendar → confirms & reminds · conflicts go to you
How it works
// how it works/01
Audit
We sit with you and map where the hours actually go. You get a ranked list of what's automatable — free.
/02
Prototype
A working agent on your real data in 2–4 weeks. You watch it run before you commit to anything.
/03
Deploy
Integrated with your tools, secured in your environment, monitored in production from day one.
/04
Iterate
The system learns your edge cases and expands to the next workflow. We stay on until it runs itself.
Where agents earn their keep
// examplesHair salon
A calendar that fills itself
Booking requests from calls, DMs, and email get answered, confirmed, and reminded. No-shows drop, evenings stay yours.
$ ncode run booking-agent
✓ 14 bookings confirmed · 9 reminders sent
Car detailing
Quotes out before the competition wakes up
Every enquiry gets a priced quote drafted from the photos and your price list — ready for your OK minutes after it lands, not days.
$ ncode run quote-agent
✓ 11 quotes sent · 3 jobs booked
Electronics resale
From buy-in to listed, automatically
New stock gets identified, price-checked against the market, and listed with photos and description — while you're at the repair bench.
$ ncode run listing-agent
✓ 23 items listed · 4 prices updated
plays well with your tools · Gmail ✦ Google Calendar ✦ WhatsApp ✦ Instagram ✦ Booksy ✦ Allegro ✦ OLX ✦ Google Sheets ✦ Stripe ✦ QuickBooks ✦ Slack ✦ Notion ✦ plays well with your tools · Gmail ✦ Google Calendar ✦ WhatsApp ✦ Instagram ✦ Booksy ✦ Allegro ✦ OLX ✦ Google Sheets ✦ Stripe ✦ QuickBooks ✦ Slack ✦ Notion ✦
// do the math
What's the boring work costing you?
Drag the sliders. We assume agents take roughly 70% of the load — conservative for well-scoped workflows.
Hours per week on repetitive work40 h
Average hourly cost$45 /h
Hours back per year
1,456 h
Cost recovered per year
$65,520
≈ full-time roles worth of time0.8
Built to be trusted
// guardrailsHuman approval
You approve what matters
Agents draft and do; you sign off at the checkpoints you define. Nothing irreversible happens alone.
Your cloud
Runs where you say
Deploy in your environment or ours, with least-privilege access to every system the agent touches.
Audit log
Every action, logged
A replayable trail of every step, tool call, and decision — so you always know what happened and why.
No training
Your data stays yours
Your data is never used to train models. Contractually, not just as a promise.
Questions
// faq// no slide decks — just a working prototype
See an agent run on your own workflow.
Tell us what you (or your team) do by hand every week. We'll tell you what an agent could take off your plate — and prove it with a prototype.
hello@ncodelab.com