Lloyds Banking Group migrated to Vertex AI in 2024 and it now has over 300 data scientists and AI developers across the Group using the platform. Unplanned platform downtime is now zero, while managed Google Cloud services allow its data scientists to supercharge their experimenting. In the space of just six months, the bank has developed 80 new ML experiments on the platform.

For Charlotte Nickels, Head of Transaction Classification Data at Lloyds Banking Group, the difference is transformative. “Our previous platform didn’t allow us to experiment at scale,” she explains. “Now, with Vertex AI, we are not only running more experiments, but quickly getting other data scientists on board to do the same. It’s really helping us release more value for customers and the bank with machine learning.”

In moving to the new platform, Lloyds Banking Group migrated 15 modelling systems, comprising hundreds of individual models from the Group’s on-premise infrastructure. As a result, the Group has saved 27 Co2 tonnes of operational emissions.

The migration has also unlocked new capabilities and tooling across the Group’s business areas, allowing it to further develop ML and AI capability and deploy advanced AI models. This includes an algorithm which reduces the income verification step in its customers’ mortgage applications from days to seconds.

Ranil Boteju, Group Chief Data and Analytics Officer at Lloyds Banking Group said, “Moving to Vertex AI has been transformative for us as a Group, providing us with the scalability and reliability to innovate with AI at pace. Vertex AI is enabling data scientists and AI developers across the Group to access GenAI solutions with consistent guardrails, as well as giving them the flexibility to use Large Language Models from third parties and open-source providers, as well as Google’s Gemini model.”

With the increasing speed at which AI is evolving, Lloyds Banking Group recognized the need for a platform capable of keeping up with these advancements. It sought a flexible solution that could integrate a variety of tools, including open-source models.

Since deploying the new platform on Vertex AI, the Group has successfully initiated over 80 new ML use cases and launched over 18 GenAI systems into production, spanning its entire business. These use cases have improved process times and the overall experience for the Group’s customers – from pricing to transaction classifications. A further 12 GenAI systems are expected to go live by the end of June.

Work is also ongoing on a cutting-edge Agentic AI system which will transform the ways customers interact with the Bank and provide a more personalized experience for end users. The prototype was built in collaboration with Google Cloud and is expected to launch to customers later this year.

“Before Vertex AI, we were limited on the number of people with access to the machine learning platform which held us back, because we need multiple data scientists if we want to operate at scale,” explains Nickels. “Vertex AI has a much lower cost of entry, allowing us to build prototypes, while giving us the scalability to deliver new use cases at pace.”