Bold Group
What everyone treated as fixed, and where the answer really was.
Engagements across retail, maintenance, healthcare, finance and manufacturing. In each, the usual fix pointed one way and the constraint sat somewhere else. Clients are not named, and every case shows how strong its evidence is.
The cases
Where the constraint really was.
Newest first. Each case shows how strong its evidence is.
How we label evidence 4 levels
- Client-measuredThe client measured the result in its own operation or accounts. The figure is the client’s own; we did not measure it.Cases 01, 02, 06
- Client-implementedThe client runs it. No figure is claimed.Case 05
- Client decisionThe client took a decision on our findings. Its operating effect has not been measured.Case 07
- DesignedWe designed it with the client. Implementation is not yet shown, so no operating result is claimed.Cases 03, 04, 08
Case 08Public registrationDesignedThe fix needed two AI tools, not an AI system.
The usual fix:
An end-to-end AI system. Where it really was: Incomplete files allowed into the queue.A design · target 12 months to 7 days, not yet implemented
Case 07ManufacturingClient decisionThe backlog lived in procurement’s buying rule.
The usual fix:
A bigger warehouse. Where it really was: Volume-discount lot sizes, beyond what the plan needed.Planned warehouse expansion cancelled by the CEO · effect not measured
Case 06Automotive partsClient-measuredInspecting plastics before the colour goes on.
Where inspection stood:
At the end of the line, on finished parts. Where it really was: Before colouring, the step that cannot be undone.Waste cost down from ฿180M to ฿45M a year · client’s own accounting
Case 05Building productsClient-implemented (NSTDA / I-TAP grant)The staircase was not what was failing.
The usual fix:
Replace the panels. Where it really was: The period between installation and handover.Implemented by the client under an NSTDA / I-TAP grant · no figure claimed
Case 04Finance operationsDesignedThe bank was already producing the record they planned to build.
The usual fix:
A claims app that captures receipts more accurately. Where it really was: The record already existed, at the bank.A design · not yet shown to be implemented
Case 03Public healthcareDesignedThe answer was already in the ward.
The usual fix:
GPS wristbands with mobile SIMs. Where it really was: Where the patient waits, and who hands him over.A design · not yet shown to be implemented
Case 02Industrial maintenanceClient-measuredThe overhaul window was a schedule, not a property of the machine.
The usual fix:
More crews, or shorter jobs. Where it really was: The order of the work: independent steps ran in series.Overhaul cycle cut from 7 days to 24 hours · measured by the client
Case 01Retail expansionClient-measuredThe stores were slow to open because of the order, not the work.
The usual fix:
Make each stage faster. Where it really was: An inherited sequence: steps waited that did not have to.~67% shorter opening cycle · measured by the client
Real client engagements, published without naming the client, to protect each client’s confidentiality and intellectual property. Numbering follows the order the cases were published.
Start here
Bring the result you need, and the thing you have been told cannot change.
One hour with a founder. You leave with a first view of where the answer may sit, and which engagement, if any, fits.
Teerasak WongpiyaCo-Founder and Managing Director
1 hour, in Thai or English, online or at your office.