Stop building data products. Start building data services.

2026-06-11 · Source: Databricks · Field: Business & Management — Corporate Strategy & Leadership, Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, medium

Summary

Howden, a global insurance broker with 25,000 employees, has transitioned its enterprise data strategy from a product-centric model to a data services approach, driven by rapid acquisitions (over one business per week last year) and the emerging demands of AI agents. Barry Panayi, Howden's Group CDO, details how their Databricks-powered platform, unifying over 100 data sources, addresses challenges like six-month integration times post-acquisition and manual reconciliation of up to four versions of the same data point. Key architectural shifts include moving data mastering and quality checks closer to ingestion and implementing a standard Accord data model to codify logic. This enables faster insight delivery, reducing "insight lag" for brokers, and leverages conversational analytics via Databricks Genie, which has saved hundreds of hours in dashboard creation by allowing direct data querying.

Key takeaway

For AI Architects designing enterprise data platforms, recognize that traditional data product models will constrain future AI agentic workflows. Your strategy must prioritize open, governed data services and shift data quality upstream to avoid perpetual reconciliation costs. Design for the anticipated pace of business growth and AI consumption, not just current needs, to ensure your architecture remains adaptable and delivers timely insights.

Key insights

Traditional data product models hinder rapid growth and AI agent consumption; shift to open, governed data services.

Principles

• Design data architecture for future pace.
• Codify data taxonomy to prevent reconciliation debt.
• Prioritize reducing insight lag over data freshness.

Method

Shift data mastering and quality checks to ingestion. Implement a standard data model to codify reconciliation. Integrate process and agentic work leaders early.

In practice

• Leverage conversational analytics for direct data querying.
• Standardize pipelines and reusable data assets.
• Unify business context across diverse data sources.

Topics

• Data Services
• AI Agents
• Data Platform Architecture
• Databricks
• Conversational Analytics
• Data Integration

Best for: CTO, Executive, AI Product Manager, Director of AI/ML, AI Architect, VP of Engineering/Data

Related on AIssential

• Architecting Data And AI In The Era Of Enterprise Intelligence: Meet Shylaja Nathan, Principal Analyst — Agentic AI demands modern data architectures that provide context, responsiveness, and semantic consistency at scale.
• Rebuilding the data stack for AI — Effective enterprise AI hinges on unified, governed, and accessible data infrastructure, moving beyond fragmented legacy systems.
• From Models to Momentum: Uniting Architects and Engineers with ER/Studio — Enterprise data modeling provides a semantic backbone, crucial for data reliability, governance, and effective AI integration.
• Your Data Layer Used to Hide Behind Your Product. Now It Is the Product. With Firebolt’s CEO — The data layer is shifting from a hidden backend to the primary product interface, driven by agent-centric interactions and deployment demands.
• AI Agents and the Fight for Customer Data — Centralized, open data infrastructure is crucial for AI agents, despite vendor data access restrictions.
• The End of Software’s Monopoly on Work — AI agents shift enterprise work from applications to a trusted data context layer, demanding unified, governed data for reliable action.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.