Evidence

Every number, labelled.

What each figure is, where it came from, and its limits.

What a project is

Status

Live implementation

Running for a customer or as a public product.

Prototype

Working build on sample or public data. Not yet in customer production.

Illustrative scenario

A worked example of how the pattern applies. Not a delivered project.

What a number is

Evidence labels

Measured

We measured it ourselves, with a repeatable method.

Reported

Reported by the customer or the delivery team. Not independently re-measured.

Targeted

What the solution is designed to achieve. To be measured in a proof of concept.

Illustrative

A planning calculation on stated assumptions. Not a result.

Evidence register

By solution

Order Intake Automation

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
60 → 6 minorder entry time per orderReportedDelivery team and customer, production use at a large enterprise (name withheld).Timed per order before and after; not independently audited.
~90%of preparation work handled by agentsReportedDelivery team, production use.Share of steps prepared without manual rework, as reported.
13 of 13real sample orders read correctly, 154 fields checkedMeasuredOur test set of real customer sample orders.Small sample; measures reading accuracy, not end-to-end time.
$243K / yearlabour saved at 500 orders a monthIllustrativePlanning calculation: 54 minutes saved per order at $45 an hour.Your volume and labour cost will differ. Labour only.

Sales Knowledge Assistant

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
99%answer accuracy across technical and support questionsReportedReference customer in industrial products (name withheld).Evaluation method and question set to be confirmed before external use.
Hundredsof products and 30+ services coveredReportedReference customer scope.Scope description, not a performance metric.
Hours → secondstime to an answerTargetedDesign goal, consistent with customer feedback.Not timed in a controlled study.

Contract & Document Intelligence

PrototypeSolution
FigureWhat it measuresLabelSourceLimits
128 of 128fields read correctly across 5 test formsMeasuredOur evaluation harness on 5 fictional sample forms (typed, scanned, handwritten), OCR plus rules.Fictional samples, small set. A customer pilot measures on real forms.
0wrong values auto-acceptedMeasuredSame evaluation.The key safety metric. Must stay at 0 on customer data.
84%of fields auto-accepted, 14% sent to a personMeasuredSame evaluation.Higher on clean typed forms, lower on poor handwriting.
−85%handling time per formTargetedDesign goal for a European payments provider back office (name withheld).Projected from 15 minutes by hand to about 2 to check. Not yet measured in production.

Customer Support Representative (CSR)

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
24/7availability, multiple voices and languagesReportedLive product capability.Capability, not a business outcome.
Full intakeinstead of empty forms, for a global consumer-goods companyReportedReference customer (name withheld).Qualitative outcome; no volume figures cleared for sharing.
Livein use at a workwear brand, a multi-clinic healthcare centre and a neurology practiceReportedDelivery team.Names withheld.

Knowledge Base

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
Every answerlinks to its source passageMeasuredProduct behaviour, visible in the live demo.Citation presence, not answer correctness.
99.2%answer accuracyReportedProduct claim.Method and test set to be confirmed before external use.

Agent Foundry: Design to Code

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
Days → hoursper design documentTargetedDesign goal for a global consumer-goods manufacturer's integration team (name withheld).Not yet measured end to end. Confirm with the delivery lead before quoting.
100%of generated documents validated against standardsMeasuredBuilt into the pipeline: every output runs through its validator.Validation coverage, not document quality.
3 + 3generator and validator agents in useReportedPlatform scope.Capability description.

Demand Intake

Live implementationSolution
FigureWhat it measuresLabelSourceLimits
0invented contacts or interface IDs, by designMeasuredCode rules: a value is kept only when the ticket evidence matches; otherwise TBD.A design guarantee, not a time saving.
Days → minutesto a first requirements draftTargetedDesign goal.To be measured with the customer's demand team.

Projected or targeted figures are never shown as customer results. Client names are withheld; sector descriptions are used with permission.

Ready to innovate, transform and thrive?

Talk to an expert 24/7. The first discovery call is free.

Book a free AI auditTry a demo