Demonstration · not a client result
AI agents and workflow automation what the deliverable looks like
The structure of the deliverable for AI agents and workflow automation, how it is produced and how you accept it, with a live specimen generated from the matching free tool’s worked example. Synthetic or permitted data only.
What this page isDemonstration
- Practice
- 09 Data, analytics & automation
- The deliverable
- 4 sectionsProduced in 5 stages, each with a review point
- Data
- Synthetic or permittedNot a client result
Demonstration of methodSynthetic or permitted data.Not a client result.
The deliverable, section by section.
What you receive, and what each part is for.
Process map
Steps, inputs, systems and failure points.
DemonstrationWorking automation
For the agreed task, with human checkpoints where needed.
DemonstrationRun logs and alerts
What ran, what failed and who is told.
DemonstrationHandover
Runbook, access and training.
DemonstrationA live specimen.
Generated from the worked example in Automation ROI calculator: synthetic data, real method. Change the inputs in the tool to see your own.
Demonstration · synthetic data · not a client result
| Hours a month | Hours | Of gross |
|---|---|---|
| By hand today: gross capacity | 50 h | 100% |
| Review: stays with a person | 6 h | 12% |
| Exceptions: stay with a person | 4.8 h | 10% |
| Maintenance: stays with a person | 4 h | 8% |
| Handed back: net capacity | 35.2 h | 70% |
- Automated, after all costs
- By hand todayUSD 900
- Automated (review, exceptions, maintenance, running)USD 326
- Minutes by handUSD 4,366 to USD 15,166
- Runs a monthUSD 5,533 to USD 14,000
- Staff costUSD 5,965 to USD 13,568
- Build costUSD 10,766 to USD 8,766
- Review timeUSD 10,414 to USD 9,118
- Exception rateUSD 10,285 to USD 9,248
- Running costsUSD 10,126 to USD 9,406
Line: base case, USD 9,766.
What it means
- Strength: Break-even in about 7.0 months.
- Note: Time saved is only value if it is used.Decide in advance what the freed hours will be spent on.
How it is produced, and accepted.
Each stage ends in something you review; acceptance is tested against criteria agreed at the start.
- 01
Inputs
List the inputs, systems and access.
Review pointInputs listed, with owners and access confirmed.
- 02
Validation
Validate the process and its edge cases.
Review pointData validated, with exceptions logged.
- 03
Analysis/build
Build the automation with error handling.
Review pointAnalysis or build checked against the agreed definitions.
- 04
Acceptance
Test with real cases and accept.
Review pointAcceptance tested by the people who will use it.
- 05
Operation
Operate with monitoring and a review date.
Review pointOperation handed over with a runbook and a refresh plan.
How proof is labelled
- DemonstrationsDemonstrations use synthetic or permitted data.
- Documented experienceDocumented experience is attributed to the organisation that held the work.
- Measured resultsMeasured results are added only once they have been measured; none are shown yet.
What needs to move forward?
Tell us the decision, the challenge or the opportunity. We reply with a scoped approach, a named principal and a fee before any work begins.
No charge to submit an enquiry.
