Live client deployments

AI workflow case studies and client-reported results.

Named, ongoing deployments showing how Mach Lilies AI Helpers support repetitive inbox, query and payment workflows while people retain control.

1-800Accountant inDinero

The deployments

1-800Accountant

From unread query to review-ready response.

Mach Lilies AI Helpers monitor unread messages in 1-800Accountant’s Queries workflow, recognise account-validation and missing-information requests, and prepare draft responses for the client team.

  • Ongoing live pilot
  • Live since
  • Human controlled
  • 67% less Reported time spent Internal pre/post staff comparison
  • $666K Estimated annualised gross staff-capacity value Client labour-hour model
  • 1 inbox Existing Queries workflow Human-controlled drafts
Adjacent live deployment · client-reported · unaudited · source detail not held

Client-reported and not independently audited; measurement period, sample detail, source detail not held, and calculation formula not supplied. This adjacent live deployment is not Quiet Quarter, not a UK-practice result, and not MTD proof. The $666K figure is estimated annualised gross staff-capacity value, not an independently audited cash saving; it is not cash saved, payroll reduction or revenue.

How the 1-800Accountant workflow runs, step by step

  1. Unread query
  2. Identify the account or missing-information issue
  3. Prepare draft
  4. Human review People decide

The Helpers prepare drafts; people retain control of client-facing decisions.

See the 1-800Accountant story →

inDinero

Two AI Helpers coordinate the payment chase.

Two Mach Lilies AI Helpers support agent-payment documentation and follow-up through a shared Outlook workflow and custom dashboard; people control financial validation, approval and release.

  • Ongoing live pilot
  • Live since
  • Human controlled
  • 47% less Human time spent Across the measured workflow
  • 82% shorter End-to-end payment cycle Across the full payment loop
  • 1,780 Reported global agent network Network population · not completed cases
Adjacent live deployment · client-reported · unaudited · source detail not held

Client-reported and not independently audited; measurement period, sample detail, source detail not held, and calculation formula not supplied. This adjacent live deployment is not Quiet Quarter, not a UK-practice result, and not MTD proof. The 1,780 figure is reported network population, not cases; the 21-person team and two Helpers are wider-process scope, not replacement scope, and financial validation, approval and release remain human-controlled.

How the inDinero workflow runs, step by step

  1. Agent query
  2. Identify missing information
  3. Coordinate follow-up
  4. Validate or route an exception People decide
  5. Continue through the wider payment process

Financial validation, exceptions and payment decisions remain with people.

See the inDinero story →

Methodology

Read the result and how it was described.

These are client-reported internal comparisons from ongoing proof-of-concept deployments. They are not independently audited, do not guarantee the same result in another workflow and are presented with their measurement basis beside the claim.

1-800Accountant
Client-reported internal pre/post staff comparison of time spent on the measured Queries workflow, with the capacity value estimated from the client’s own labour-hour model.
inDinero
Client-reported internal comparison of human time and end-to-end cycle length across the measured payment workflow during the ongoing deployment.
What this shows

An adjacent operating pattern, demonstrated.

These client-reported deployments show a controlled inbox operating pattern in adjacent live workflows. They are not presented as MTD or Quiet Quarter performance results.

See the related operating pattern illustrated for UK accountancy work: watch the complete Monday-to-Friday demonstration (a simulation with sample data) or read how Quiet Quarter is scoped for accountancy practices.

The next useful step

Bring us one chase loop.

We will tell you whether it is worth installing, what can move safely and what must stay human-led.

Book an AI workflow assessment Take the readiness scorecard