Book a scoping call →
Solutions by outcomeNever miss work26Every call, lead and enquiry answered, routed or captured.Get paid faster32Money collected, matched and reconciled without chasing.Fill the diary12Fewer no-shows, fuller weeks, less phone tag.Answer without staff16The repeated questions handled, the rest escalated.Join up the systems77Data entered once, moving everywhere it is needed.Verify and secure15Real customers in, fraud and fakes out.Prove you are compliant8Obligations met by default, evidence captured automatically.Run it for me12Someone watching, someone answering, every month.
Solutions by capabilitySMS & messaging28Voice & call flows22Verification & identity13WhatsApp & chat5Video & telehealth4Email & notifications4Flow automation3Unified conversations2Contact centre5Delivery & tooling2Compliance & records5Workflow automation49Trade & job management8Applied AI8Self-hosted platforms7Payments & billing29Run it for me4
All 198 builds Describe it in plain English Venture models Instant scope & price Free tools Search everything Connections Integration guides Workflow library By integration Partnerships Contract roles Honest comparisons Field notes Work we have built Colabs How it works Platforms & licensing Pricing Book a scoping call
Home / Work / Ledgerloop — AI Back Office for Trades
Business software Colabs ventureIn production

Ledgerloop — AI Back Office for Trades

Overdue invoices chase themselves, in the business owner's name, and payments reconcile on arrival.

What we set out to build

To remove the single most-avoided job in a trade business: chasing money. The founder wanted it to happen without the owner having to be the bad guy, and without touching their funds.

What was built

  • Automated invoice chasing that reads as the business, not as a debt collector
  • Reconciliation the moment a payment lands
  • A job-management wedge: connect the field service platform and the accounting system, and the loop closes itself
  • A seven-phase AI account-manager agent loop, live in production
  • Read-only access by design, so the product can never move a customer's money

What it achieves

  • Runs offline against mocks for development, so a demo never depends on a client's live data
  • The safety invariants of the agent loop are documented as non-negotiable rather than left to code review
  • Read-only is a structural guarantee, not a policy: the integration scopes do not permit movement of funds
The hard part: An AI agent that emails your customers about money is one bad output away from a lost client. The engineering went into constraint: what it may say, when it must escalate, and what it can never do.

The stack

TypeScript, Xero, ServiceM8, React.

Common questions

How was this built?

This was built through Colabs, our sister company. A founder brought the idea and the domain knowledge, Colabs brought the team, the founder funded the build with a monthly subscription, and they hold equity in the company that resulted. We label every case study, because a studio that publishes its prices and argues the other side of its own comparisons has to be equally precise about its own portfolio.

What state is it in?

In production. We state that plainly rather than describing everything as production: several things here are deliberately gated, and the reason is usually evidence rather than a missing feature.

Can you build something like this for us?

That is the point of publishing it. The same patterns are in the catalogue at published prices, and you can scope your own version on the plain-English page without contacting us first.

Where to next

If something here is close to what you need, the same patterns are in the catalogue at published prices, or you can describe your version in plain English and get it scoped and priced without talking to anyone.

[ // / BUSINESS SOFTWARE ]

Other business software work

Built with the same discipline, in the same domain.

[ COLABS VENTURE ]

ChatVault — Document Chatbot Generator

Turn a business's own documents into an embeddable chatbot that cites its sources.

Built →
[ COLABS VENTURE ]

LearnLedger — Local-First AI Bookkeeping

Bookkeeping assistance that runs on the user's own machine rather than in someone else's cloud.

Pre-launch →
[ BETR INTERNAL ]

This Site and Its Quoting Engine

842 pages, a plain-English quoting engine, and a validator that fails the build on thin content.

In production →

Want something like this built?

The catalogue prices the patterns behind this work. Describe your version and the engine will scope it, including an allowance for anything nobody has costed.

Describe it in plain English → See all work