The Signal
For most of the last decade, enterprise analytics followed a BI replacement cycle: older reporting tools replaced by newer ones, data warehouses modernised, dashboards rebuilt. The DIA layer, the intelligence and analytics component of the Digital Business Platform (DBP) architecture in D3, was treated as a reporting function inside a broader platform build. That framing is now breaking down. Composable data platforms are being adopted as DIA foundation architecture, not analytics tooling.
Databricks, Snowflake, and cloud-native data mesh patterns are moving the procurement conversation from the analytics team to the platform architect. AI and ML orchestration costs on cloud-native platforms dropped 40-60% in 2025-2026, removing the cost barrier that previously kept DIA tightly coupled to legacy BI tooling. Major enterprises in financial services, retail, and logistics are implementing data mesh architectures as their DIA foundation layer. Gartner and Forrester both report that DIA budget ownership is moving to a dedicated platform architecture function in organisations above 5,000 employees.
“The DIA-within-BI framing has broken down, the procurement conversation is moving from the analytics team to the platform architect.”
The Direction
Within 6-18 months, the DIA platform layer will be a recognised, named architecture component in enterprise technology strategy at the level of ERP and CRM. Organisations that have already separated it will be positioned to compose AI capabilities on a purpose-built foundation. Those still treating it as a BI function will face a reclassification and rebuild cost when the gap becomes operationally visible.
“The gap compounds with every AI capability decision locked in on the wrong foundation.”
The Implication
Practitioners who continue to classify the DIA layer as a BI function will build AI and ML capabilities on a foundation not designed to support them. Model quality, data latency, and governance accountability will all carry the cost of that misclassification, and the gap compounds with every AI capability decision locked in on the wrong foundation. Audit your current DIA architecture classification this quarter. If it sits inside a BI programme or reporting workstream, raise the reclassification question with your platform architect and CDO before your next AI capability decision locks in the wrong infrastructure.
Sources
- 01Databricks, Snowflake, cloud-native data mesh adoption reporting (2025-2026)
- 02Gartner / Forrester enterprise survey: DIA budget ownership shift to platform architecture function
- 03AI/ML orchestration cost reporting: 40-60% decline on cloud-native platforms (2025-2026)
