Our approach
One method,
from advice to a running system.
Discover, design, build, evaluate. We start from how the work runs and write the specification before any code, then build to it and test it. Most multi-entity groups start with the AI Transformational Audit.
Response specification
- Trigger: temperature alert
- Action: check rear door
- Proof: driver confirmation
Specifiedbeforecode
Truck 12
Temperature alert → Driver action
Door check confirmed
Illustrative scenario · Rasikh Logistics cold chain
Response specification
- Trigger: temperature alert
- Action: check rear door
- Proof: driver confirmation
Specifiedbeforecode
Truck 12
Temperature alert → Driver action
Door check confirmed
Illustrative scenario · Rasikh Logistics cold chain
- 01
Discover
Understand the operation
- 02
Design
Agree the specification before any code
- 03
Build
Implement the response
- 04
Evaluate
Verify the agreed behaviour
Ways in
Start from your situation.
Pick the situation that sounds like yours.
Each path starts with one service and can continue to a build.
For multi-branch and multi-entity groups, we recommend starting with the AI Transformational Audit.
“We don’t know where AI fits.”
Discover
Find where AI pays off, with evidence. The Audit delivers an Executive Report, maps of where the work waits, what to build first, ranked, and the expected return, with every assumption written down.
Recommended start for multi-branch and multi-entity groups
AI Transformational Audit- AI Readiness Assessment
“We know the problem, not the spec.”
Design
The AI Solution Blueprint: the workflow, architecture, integrations, acceptance criteria and a defined implementation quote.
“We have an approved Blueprint.”
Build
We build your approved Blueprint into your systems, test it against its acceptance criteria and hand it over with a runbook.
“Our AI is live. Is it working?”
Evaluate
An independent evaluation of a live or vendor-built system against agreed criteria, with the evidence and what to fix first.
The one required gate
No spec,
no build.
We build only from a paid, approved AI Solution Blueprint. It fixes the scope, the integrations and the acceptance tests before the first line of code.
AI Solution Blueprint
Our principles
Four rules
we work by.
They apply whether we advise, build or evaluate.
Illustrative example · supplier onboarding, Rasikh Holding Group
- 01
Operations map
- Company request
- Procurement
- LegalWaits here
- Group finance
Begin with operations
We map how the work really runs across your companies, systems and approvals before we suggest any AI.
- 02
Decision owner
Group procurement director
New suppliers above the group limit stay with a person.
Signed offPeople own the decisions
We design where people approve, review and answer for the outcome, and let AI carry the work around them.
- 03
Evidence log
ExpectedObserved- Duplicate supplier flaggedFlagged
- Arabic CR document readRead
- Missing VAT certificate sent backMissed, fixed, retested
Verify with evidence
We test against agreed acceptance criteria and your real cases, and show you what was expected against what we observed.
- 04
Handover sheet
- Owner
- Procurement operations
- Escalation
- Group procurement director
- Rollback
- Last approved version
Runbook attached
Hand over with clarity
Every system is handed over with its runbook, a named owner and an escalation path.
Your next step
Not sure where
to begin?
Begin with the Audit: it shows where AI belongs in your operation. Already know your stage, whether discover, design, build or evaluate? Talk to us and we pick up from there.