Tooling & infrastructure
The maturity of the AI/ML stack, vendor strategy, integration capability, and platform discipline.
What we ask
The four questions in this dimension, with the four ordinal options and their fixed scores. Identical wording to what you will see in the assessment.
1. How is your organisation provisioning AI capabilities (models, APIs, agents) today?
A coherent stack saves cost, simplifies governance, and improves reliability. Fragmented vendor sprawl creates compliance and integration debt.
| Option | Response | Score |
|---|---|---|
| a | Consolidated platform with vetted vendors; usage governed centrally | 100 |
| b | Two or three preferred vendors, lightweight governance | 67 |
| c | Each team picks their own tools | 33 |
| d | Mostly shadow IT — we don't really know who is using what | 0 |
2. Are your AI integrations capable of connecting to your existing systems of record?
AI value comes from acting on real business data — CRM, ERP, support, billing. Integration capability is often the bottleneck.
| Option | Response | Score |
|---|---|---|
| a | Production-grade integrations across major systems; new ones ship in days | 100 |
| b | Integrations exist for top 1–2 systems; others manual | 67 |
| c | Mostly manual data movement / copy-paste | 33 |
| d | AI tools live in isolation from business data | 0 |
3. How do you monitor AI model performance in production?
Models drift. Without monitoring you find out via customer complaint or regulatory inquiry, not your own dashboard.
| Option | Response | Score |
|---|---|---|
| a | Continuous monitoring with alerting on drift, latency, error rate | 100 |
| b | Periodic manual reviews | 67 |
| c | Ad-hoc, only when something visibly breaks | 33 |
| d | No monitoring | 0 |
4. How quickly could you switch AI providers for a critical use case if needed (price hike, outage, compliance change)?
Vendor lock-in is a real risk in 2026 as the AI provider landscape consolidates and pricing models change.
| Option | Response | Score |
|---|---|---|
| a | Provider abstraction in place; switching is days of engineering | 100 |
| b | Switching possible but would take weeks of rework | 67 |
| c | Significant rebuild — months of work | 33 |
| d | Effectively locked in | 0 |
Ready to see your full score?
Take the full assessment to see your score across all six dimensions, your peer benchmark, and your three highest-leverage moves.
Take the assessment →