Find the decision
Identify the point where AI could change the result, the cost or the risk.
Software engineering and testing
AI speeds up the build. Make the testing keep up.
Trueform works with engineering and test teams to put AI into the whole delivery lifecycle, with testing called out rather than bolted on. We find where assurance is holding the release back, build the evidence into delivery, and keep it running as the code and the models change.
f.01
When AI speeds up the build, testing becomes the constraint.
Code, tests and documentation now arrive faster than most teams can check them. Outputs vary, models change underneath the product, and a green demo is not evidence that a system is ready for real work.
Trueform works across the full lifecycle, from requirement to release, and calls testing out explicitly: what the system must prove, how that evidence is produced, and how it is rerun when anything changes.
f.02
Put the controls into the system.
Choose the model
Use frontier reasoning only where the work earns its extra cost.
Set the boundary
Decide what the model may see and change, and what comes back to a person.
Measure the result
Agree the evidence before the build, then record time, cost, quality and risk as the system changes.
f.03
Find the point where AI starts breaking the work.
Start with one process, product or system. Find where AI is moving faster than the organisation can control it or prove what it is worth.
If there is no useful boundary to improve, we say so. You leave with a decision about what should happen next.
f.04
The senior named in the work stays in it.
Trueform stays small. The senior responsible for the engagement, and any specialist named in the work, remain close from diagnosis through adoption.
The practice is based in the UAE and works across the GCC. You know who is making the judgement call and who remains accountable when the work changes.
About the practicewhat we do
Three ways we work.
Software Engineering & Testing Efficiency
For engineering and test teams building with AI. We put AI into development and testing together (test strategy, automation and the evidence a release needs), so assurance keeps pace with the build instead of holding it up.
Internal Training Packages
Practical AI capability for your own people, run as an internal programme. Our lead package is Trueform AI Essentials.
About AI EssentialsGeneral Process Consultancy
For operational work outside software delivery. We find the process where AI would change the result, decide whether it belongs there, and build or recover the change with the evidence to trust it.
f.05
How to engage
Start with the smallest piece of work that can settle the decision. Build, recover or stay involved only when the evidence holds.
Readiness Diagnostic
A 30–45 minute conversation about one process, product, system or AI pilot. We identify where AI could help, what would make it unsafe or uneconomic, and whether there is enough value to justify a next step.
AI Readiness Sprint
A written assessment of the value case, delivery boundary and production risks. The full fee is credited against a Foothold started within 60 days.
Foothold Engagement
One end-to-end AI-enabled change, with the evidence agreed before the build, something tangible to review and a decision each week.
Strategist Retainer
Ongoing access to the same senior after the first proof, including review when models, prompts, tools or permissions change.
f.06
Where Trueform will say no
There is no useful result to improve.
The work depends on a vendor the client cannot leave, and that risk has not been accepted.
The engagement would end in slides rather than a working change.
Nobody owns the value or the spend.
f.07
Current Work
In progress
Clinical capture, at the point of delivery
Voice-first, offline-capable clinical capture designed for the stable, so the vet can record the work without leaving the horse. The application is being prepared for its first stable pilot.
case study
AI Essentials at a UK consultancy
Write-up in progress
A UK consultancy ran an AI foundations programme for its team in September 2026, with a baseline assessment before and after. The before-and-after results are being written up and will appear here.
Vol 1, no. 3 · Trueform · 11 September 2026
Agents Outside Their Scope
They were set tasks, they found the shortest route to completing them, and the routes ran through places nobody was watching.
Read the notef.09
Bring the piece of work you are unsure AI should touch.
The first conversation decides whether there is useful work here. If there is, we will recommend the smallest next step. If there isn't, we will say that too.