Challenge

Accenture worked with a banking customer to build a Fostering Better Finance application. The project aimed to identify vulnerable people from their credit and debit transaction profiles and intervene early to drive better financial outcomes. This project relied heavily on using their customers' most sensitive transaction data.

• High barrier to work with partner
  Accenture's banking client was excited to run the project, however saw major security and governance hurdles to jump over in order to grant data access.
• High level of risk
  The team recognised the risk of working directly with this data and the potential damage to their business it could cause if they were responsible for a data breach.
• Long lead time to work with customer data
  Accenture and their customers want value delivered quickly. It's part of Accenture's competitive advantage. Long lead times to get projects started can dampen the momentum of great initiatives.
  

Solution

Patrick Hegarty at Accenture leveraged synthetic data to get his project started, build and test prototype applications and provide demonstrable proof to the client that they could deliver value. This was achieved with zero risk to any real customer data.

“Using Hazy's banking transaction generator, Hazy generated a representative synthetic data set that contained detailed account information about fictional individuals based on the client's data content. These individuals also had associated transaction profiles covering three years of interactions and account usage.

This data contained no real customer information but generated realistic customer profiles and behaviours. This meant Patrick could immediately begin to build the Fostering Better Finance application at no risk.”

“Accenture deployed the Hazy Generator on-premise within their own cloud based development environment. This meant they didn't have to deploy any part of their technology stack on to their customers premise to get the project started.

Multiple users at Accenture used the Generator to create synthetic data sets when they needed it to test and develop the Fostering Better Finance application. This approach was a lot more flexible than the heavy user restrictions that are usually placed on real data.”

Impact

The Fostering Better Finance project was successfully delivered to their customer and is now being rolled out to other clients. Hazy synthetic data underpinned the success of the initial project and can be used in the future to help other clients to use the application.