Work: AI Transformation Inside Real SMEs | ProcesoAI
Skip to content

Work

What this looks like inside real businesses.

We don't name clients. The work sits too close to their operations and their numbers. So we describe the problem instead, and we label what's delivered and what's still in build.

New Zealand / Trades

Turning a founder dependent trades business into a sellable asset

The founder quoted and invoiced at night, recurring servicing revenue went untracked, and ad spend had no visibility. We're building the platform that changes that.

Garage door installation and servicing · ~20 people · founder-led

Status

In build

Method phases

DiagnoseRoadmapImplement (in progress)

The problem

The founder is the business. He quotes and invoices at night. Garage doors need servicing every 12-18 months and nobody was tracking it, so recurring revenue was walking out the door. Google Ads drove leads with zero visibility into cost per lead or which campaigns produced revenue. The founder is planning his exit, and the business is unsellable as it stands.

What we built

One platform for the whole operational lifecycle: booking capture, site visit scheduling, quoting, invoicing with deposit and completion charges integrated into Xero, and a marketing attribution dashboard. On top of it, an assistant that pulls from the founder's email, calendar, the app and Xero to answer any business question.

Expected impact (not yet measured)

  • Recurring servicing revenue recaptured
  • Founder's evenings back
  • Marketing spend decided on unit economics
  • A business that can be sold

Australia / Cleaning services

Quoting, invoicing, payroll and inbox automated on top of the tools she already had

The founder was doing every quote, invoice, payroll run and client reply by hand. No new software: we fixed and connected what she had and built assistants on top.

Residential cleaning · small team · founder-led

Status

Delivered September 2026

Method phases

DiagnoseRoadmapImplementEvolve

The problem

The founder was doing quoting, invoicing, payroll and every client reply by hand. Her inbox was out of control. Xero and her job management tool were misconfigured with stale data, so neither was usable.

What we built

No new software. We fixed the settings and purged stale data in Xero and the job tool, connected them to Stripe so invoices are paid by card automatically, and built assistants for the founder: quotes from client enquiries, invoicing, payroll, a weekly financial summary, and inbox triage with a draft reply for every incoming message.

What changed

  • Daily admin down dramatically
  • Inbox managed with drafted replies
  • The founder is back on sales and the team
  • Hours saved will be published once a full month is measured

A third engagement (a pharmaceutical distributor in Central America, three AI assistants in Microsoft Teams for 18 field reps) is in scoping and will appear here when it starts.

If one of these sounds familiar, start with a Scan.

A free, specific look at where AI changes your numbers. 3-5 business days, no call needed.

Get my Scan