The lesson
Lead with quantified operating results, then use the customer’s architecture change to connect data plumbing to strategic AI work.
The page opens with the NAB logo, the title “National Australia Bank enhances customer experiences and powers GenAI,” and a right-hand fact panel for company size, region and industry. A “Key results” block lists cost, model performance and user migration outcomes before a quote from Joanna Gurry frames Fivetran and Databricks as the replacement for complex legacy technology. The story then moves through legacy data problems, Project Ada, “Fivetran cuts data ingestion costs by 50%,” “Unlocking NAB’s data across the enterprise,” and “Future-ready with a modular architecture.” It names specific enterprise uses, including banker chat assistant, financial crime detection, campaign refinement and AI-led document review.
Don't copy NAB’s Project Ada name, banking use cases, Databricks on AWS setup, or Joanna Gurry’s language about legacy technology. Keep the structure of results first, customer context, old-system pain, architecture choice, named results, and concrete downstream use cases.
Tell us what you want to make. We’ll talk through how to put these ideas to work for your brand.