Explore the publication
Data & AI
Data platforms, governance and practical enterprise AI adoption.
Explore this topic
Data & AI · Analysis
AI procurement: buy a change process as well as a feature
An AI feature can change after approval. Procurement needs clear ownership of model, data-use and feature changes throughout the supplier relationship.
Data & AI · Practical guide
Build an AI evaluation set that can overturn the favourite model
Create an LLM evaluation dataset from authorised tasks, explicit scoring and held-out examples. Test consequential failures before choosing a model.
Data & AI · Explainer
Data lineage and observability: choose the evidence your investigation lacks
Compare lineage and observability through a wrong dashboard result and a late data feed. Check product coverage before choosing a platform.
Data & AI · Explainer
RAG or fine-tuning? Start with what the AI gets wrong
RAG supplies information at response time; fine-tuning adapts model behaviour through training. The right choice starts with the failure you need to fix.
Data & AI · Decision guide
AI data readiness includes the document you have just revoked
Prepare enterprise retrieval around permission changes, freshness and deletion. Test every derived copy, not just the quality of the source documents.
Data & AI · Decision guide
Enterprise AI platforms: make the shortlist prove the workflow
For enterprise AI leads, the decisive comparison is what each platform delivers inside a real workflow, with real permissions, review effort and support costs.
Data & AI · Decision guide
Buying a data catalogue? Ask it which revenue number to use
For a data platform lead, catalogue value depends on whether people can find, understand and use the right data. Connector counts leave much of that job untested.
Data & AI · Practical guide
Build an AI use-case inventory that can trigger a decision
Record the business use, data, authority and owner behind each AI deployment. Build an inventory that exposes changes and sends unresolved uses to a decision.
Data & AI · Practical guide
Data quality rules: protect the billing decision, not the dashboard score
Turn billing failure modes into precise data quality rules, with clear tolerances and actions. Completeness alone cannot establish whether a customer should be charged.
Data & AI · Practical guide
An AI acceptable-use policy should answer the employee’s next question
Write AI policy around tasks, data and review decisions. A supplier-proposal example shows how to give staff a permitted route and a useful escalation path.
Connected topics
