Suntory Oceania Enhances AI Capabilities with Databricks and Copilot Pro
Suntory Oceania is advancing its artificial intelligence (AI) initiatives through investments in Databricks and Copilot Pro, while concurrently examining native agentic options for its SAP core systems. Matt Dixon, the company's digital and IT director, shared insights on this development during the SAP NOW AI Tour in Sydney.
Over the past two years, Suntory Oceania has transitioned roughly 90% of its datasets, totaling about 40, to Databricks, establishing a centralized data foundation designed to support AI-driven applications. This shift is part of a broader effort to harness AI technologies and integrate these capabilities with their existing business processes.
Dixon explained, "We began migrating from our legacy data platforms to Databricks two years ago, and now a significant portion of our data infrastructure is leveraging this environment. This allows us to develop tools on top of Databricks and eventually integrate them into various customer experience applications."
In addition to Databricks, Suntory Oceania has been deploying Copilot Pro to enhance personal productivity across the business. Dixon noted that the company is currently acquiring a few hundred licenses for Copilot Pro, with plans to expand this number in the coming year. The company is ensuring broad access to the standard version of Copilot throughout the business.
Another focus area is agentic AI, where the company is assessing solutions that are native to its SAP systems. Dixon emphasized the importance of first exploring options within the enterprise's core SAP environment before considering external vendor solutions.
As part of its AI vision, Suntory Oceania aims to use SAP for autonomous and data-driven operational enhancements. A particular area of interest is demand sensing, which involves using near real-time data signals to adjust supply chains proactively. Currently a manual process, Dixon envisions a future where AI agents could optimize demand planning algorithms and automate tasks like raising stock transfer orders.
Dixon highlighted the evolving landscape of AI within the company, stating, "These advancements are never going to be real-time, but they will certainly become more immediate compared to the current processes. We foresee a future where agents handle these demand sensing tasks efficiently, with minimal human intervention, primarily to verify and approve generated suggestions."
