Build, buy and run

8 tools in this group, all free with a printable result

1. 
   Decision tool

   AI Build vs. Buy vs. Partner

   Should we build this ourselves, buy a platform, or partner?

   When to use it

   At the start of any AI initiative with a real budget behind it.

   What you get

   One recommendation with a confidence score and risk flags.

   Effort

   7 inputs

   Last updated

   29 Sept 2026

   Get my recommendationBackground reading: Not Every Problem Needs a Model. Some Just Need a Lookup Table
2. 
   Assessment

   AI Vendor Evaluation Scorecard

   Which of these two vendors should we sign with?

   When to use it

   At shortlist stage, before the contract goes to legal.

   What you get

   A side-by-side recommendation across nine dimensions, with risk flags.

   Effort

   2 vendors, 9 dimensions

   Last updated

   2 Oct 2026

   Compare vendorsBackground reading: Why the AI Model Is the Weakest Link in Your Supply Chain
3. 
   Checklist

   Outcome as a Service Qualification Checklist

   Is this deal a good fit for outcome-based pricing?

   When to use it

   When a vendor offers to be paid on results instead of licences.

   What you get

   A pass on 8 threshold gates, a score on 55 criteria and a recommended commercial structure.

   Effort

   63 checks across 9 domains

   Last updated

   29 Sept 2026

   Take the checklistBackground reading: Outcome as a Service: What the Contract Has to Do
4. 
   Decision tool

   Data Architecture Decision Tool

   Lakehouse, data mesh or data fabric: which one fixes our data first?

   When to use it

   Before you buy any AI model, if your data sits in many places.

   What you get

   One architecture recommendation with the reasons behind it.

   Effort

   7 inputs

   Last updated

   29 Sept 2026

   Choose your architectureBackground reading: Messy Data Is Why Most AI Projects Fail (Fix This First)
5. 
   Decision tool

   Fine-Tune vs. Retrieval-Only Classifier

   Is this dataset safe to train into a model, or should it stay behind retrieval?

   When to use it

   Before any fine-tuning project that touches internal or personal data.

   What you get

   A 0-100 suitability score and a clear fine-tune or retrieval-only call.

   Effort

   7 risk factors

   Last updated

   29 Sept 2026

   Classify your dataBackground reading: How to Build Governed RAG 2.0 Systems for High‑Stakes Use Cases
6. 
   Decision tool

   AI Incident Response Planner

   If one of our AI systems fails tomorrow, who switches it off, and which regulator clock has started?

   When to use it

   Before a system goes live, or when the board asks what happens if the chatbot gets it wrong.

   What you get

   A printable incident playbook: incident types, severity levels, a first-hour, day and week runbook, roles, notification clocks and gaps.

   Effort

   1 system, 14 questions, about 10 minutes

   Last updated

   25 Oct 2026

   Build the playbookBackground reading: What the DBS Outages Teach Boards About Technology Risk
7. 
   Dashboard

   Model Performance and Health Dashboard

   Is our deployed model drifting, slowing down or failing?

   When to use it

   Once a model is in production and someone owns keeping it healthy.

   What you get

   Drift, latency, uptime and incident tracking, with why each metric matters.

   Effort

   Adjust the inputs to your model

   Last updated

   17 Sept 2026

   Open the dashboardBackground reading: What Keeps You Awake at Night About AI?
8. 
   Dashboard

   Human-AI Interaction and Decision Quality

   Are our people over-trusting or under-using the AI?

   When to use it

   When a system is live and decision quality matters more than throughput.

   What you get

   Acceptance rates and automation bias benchmarked against McKinsey, BCG, Stanford HAI and MIT research.

   Effort

   Set your industry, role and decision type

   Last updated

   17 Sept 2026

   Open the dashboardBackground reading: AI Fatigue: Why Verification Is Harder Than Creation – And How to Work Differently