Manufacturing & Logistics
Support that answers itself. 68% fewer routine support requests
The team now gets far fewer routine requests, and more customers are helped the first time they ask.
The client
A software company that helps businesses manage deliveries and supply chains.
- Sector
- Manufacturing & Logistics
- Published
- Anonymised at the client’s request.
Case 01 / 01 · Manufacturing & Logistics
01 · Who they are
A software company that helps businesses manage deliveries and supply chains.
What Intelligent Systems Engineering involves.
- 01Agree what “working” means
- 02Build the data pipeline
- 03Choose a right-sized model
- 04Adapt the model only if it pays
- 05Test on your real cases
- 06Alert before a customer notices
Agree what “working” means
Define the task precisely enough that "working" has a measurable definition — accuracy, latency, cost per call — before a line of production code exists
Build the data pipeline
Build and maintain the data pipeline behind the system: ingestion, cleaning, labeling, and the ongoing process that keeps the data from decaying after launch
Choose a right-sized model
Select the model architecture based on the problem's actual requirements, including the frequent case where the right answer is a smaller, cheaper, more reliable model than the newest release
Adapt the model only if it pays
Fine-tune or adapt a model where that earns its cost in accuracy or latency, and skip it where a well-prompted general model already clears the bar
Test on your real cases
Build an evaluation harness that runs against real cases pulled from your own data, not a handful of examples someone picked because they looked good
Alert before a customer notices
Instrument the deployed system so a failure surfaces as an alert to an engineer, not as something a customer discovers and reports first
Questions people ask before buying
01 / 06
Does it cost anything to ask?
No. You describe one problem, plainly, in a few sentences, and we tell you if it is worth building. That answer is free. No discovery invoice, no proposal deck.
02 / 06
What if AI is the wrong tool?
We tell you. A simpler rule, or no change at all, is sometimes the honest recommendation, and we would rather say so than build something that should not exist.
03 / 06
How will we know it works?
Where an engagement carries a measurable target, that target is written into the contract with a defined test, not implied by a case study.
04 / 06
Who can see our data?
Client data is processed only for the engagement it was provided for, and only by the people working on it. Access is granted per project and removed when the engagement ends.
05 / 06
Do we own what you build?
Documentation and runbooks are written for the engineers who take it over. We would rather leave cleanly than become a dependency you cannot remove.
06 / 06
Can you guarantee the result?
No, and you should treat any firm that does with suspicion. The figures on this site are real, but each one came from one specific client. What we can do is write your target down and test against it.
Your problem
Describe it plainly. We will tell you if it is worth building.
Next case study · 15
They needed to know if the new market would buy. In four months they had 15 serious sales leads, then did the same in France.