UBS has embarked on an extensive data transformation initiative aimed at developing two specialized AI assistants, collectively referred to as UBS Red. This innovative solution leverages Azure AI Search and Azure OpenAI Service to enhance support for client advisors during client interactions, featuring multi-language support to streamline responses to client needs.
Michel Neuhaus, the head of AI, data, and analytics for Personal & Corporate Banking at UBS, highlights the complexity of the challenge: “The wealth of internal information we possess, including our financial expertise and insights into global market trends, is immense. Our goal is to make this information readily accessible to all employees at the bank, and Microsoft plays a pivotal role in facilitating that.”
In response to the vast amount of data available, UBS developed a queryable knowledge base. The implementation of Azure AI Search has enhanced information retrieval strategies and enabled the extraction of new insights from unstructured data through advanced language model reasoning. Enhanced vector search capabilities allow the organization to locate information based on conceptual similarities across diverse data types, thereby maximizing the value extracted from its extensive data resources.
Claire Corish, the head of analytics and sales platforms within the Operations and Technology Office at UBS, notes, “We have digitized around 60,000 documents related to investment advice and products. This queryable knowledge base built on Azure is now easily accessible to our staff, significantly reducing the time required for meeting preparations and research.”
The unlocking of this data repository has paved the way for UBS to establish an internal AI Hub dedicated to exploring new applications of artificial intelligence.
Looking ahead, Jonas Isliker, the head of AI, data, and analytics for Global Wealth Management at UBS, emphasizes the organization’s commitment to AI transformation: “Our primary focus is to swiftly advance our AI initiatives. While delivering flagship AI solutions like Red is vital, equally important is the development of a scalable framework that accelerates the deployment of AI use cases. This entails implementing a data mesh, migrating to the Azure cloud, establishing new AI risk governance structures, and creating AI factories to ensure rapid and agile solution delivery. These components are crucial to meeting the escalating demand for AI applications throughout the organization.”
In 2022, UBS announced ambitious plans to transition over 50% of its applications, including essential workloads, to the Microsoft Azure cloud within five years. Within the past ten months, the Azure OpenAI Service has been deployed across all major divisions of the bank globally, reaching 30,000 employees as of December 2024.