Maureen Costello, vice president UK, Ireland, Sub-Saharan Africa, Google Cloud, opened the keynote with a sweeping overview of the last 12 months of innovation.
“In 2024, we released more than 3,000 product advancements across Google Cloud and Workspace,” she told attendees. “We expanded to 42 regions, including Sweden, Mexico and South Africa, and extended our 2 million mile terrestrial and subsea fibre network with new cables.”
At the heart of this expansion is generative AI. According to Costello, the platform has seen a 40-fold increase in Gemini usage since last year, and over seven million developers are now using Vertex AI and AI Studio. AI is also assisting Google Workspace users over two billion times per month, underscoring the sheer scale of AI adoption across the company’s ecosystem.
“Today, that future is being built by all of us,” said Costello. “When I speak to customers and reflect on the pace of change, it’s faster than it’s ever been.”
Among the 400+ customer stories highlighted at the event was a collaboration with the Imperial War Museums. In partnership with Capgemini, Google Cloud’s generative AI tools were used to digitise and transcribe over 45,000 oral history recordings.
“Imagine hearing from your ancestors about their experience in war – that’s exactly what we’ve done,” Costello explained. “These transcriptions will soon be available to the public, researchers and educators around the world.”
The summit also spotlighted how AI is transforming everyday banking. UK challenger bank Starling unveiled Spending Intelligence, an AI-powered financial assistant built with Gemini on Google Cloud.
Harriet Rees, CIO of Starling Bank, joined Costello on stage to showcase the tool.
“With Spending Intelligence, powered by Gemini, customers can use natural language, text or voice to ask really simple and intuitive questions,” Rees said. “Like, ‘How much did I spend on transport last month?’ or ‘How much did I spend on holidays last year?’”
Beyond just convenience, Rees said the goal is to change customer behaviour through better insight and control: “We want customers to really get closer to their money… to understand a bit more about what they might want to change.”
Starling’s AI journey goes further than consumer tools. Through its Engine platform, the bank is already exporting its technology to institutions in Romania and Australia, and has plans to scale globally.
“This year, our one aim was to bring AI into the hands of our customers,” Rees added. “We wanted to show them that we can actually change their customer experience with the power of AI.”
On choosing Google Cloud as a long-term partner, Rees said: “Google was the first cloud that was able to provide me with assurances around data sovereignty and security. It was a natural choice.”
Costello also introduced a new AI tool called Extract, designed to digitise old planning documents such as maps and handwritten notes using Vertex AI and Gemini. The tool is expected to radically accelerate housing and infrastructure planning.
Other examples included UK grocery giant Morrisons, which is using BigQuery and Gemini to power an in-app product locator, and Pearson, who are collaborating with Google Cloud to build AI tools that adapt to students’ learning styles.
“Our collaboration will provide educators with AI-driven insights and time-saving tools,” said Costello. “This means teachers can focus on what matters most – sparking curiosity, developing critical thinking, and fostering a love of learning.”
Healthtech startups like Dietetics AI, and food sustainability firms like Marty AI, also received praise for using Google Cloud to scale their impact, demonstrating how accessible AI is becoming to early-stage innovators.
Hayette Gallot, Google’s VP of AI platforms, took the stage to break down Google Cloud’s AI stack, starting with its new AI Hypercomputer.
Powered by TPU v7, the infrastructure offers a 30x performance gain over 2018 models, enabling enterprises to train large-scale AI models faster and more cost-effectively.
“AI is reshaping everything at a speed we never imagined,” said Gallot. “It’s about optimising at every layer, infrastructure, models, platforms and agents.”
One of the key announcements was Gemini on Google Distributed Cloud, enabling customers to run AI models directly on-premises to reduce latency and meet data residency needs. Gallot also highlighted Deutsche Bank’s use of this setup for its Autobahn FX trading platform, reducing development cycles from weeks to hours.
Google also introduced a suite of agent development tools, including the Agent Builder, Agent Engine, and the Agent-to-Agent (A2A) protocol, which allows intelligent agents to interoperate securely.
“We want to ensure the user experience is coherent,” Gallot said. “If agents are going to be everywhere in the enterprise, they need to work well together.”
To support this, Google launched an Agents Marketplace, giving businesses a one-stop shop to find, buy, or customise pre-built AI agents.
Sustainability was a recurring theme. Google Cloud reaffirmed its goal of reaching net-zero emissions by 2030, and highlighted its leadership in “24/7 carbon-free energy” for its global data centres.
“AI can drive efficiencies, reduce costs and support sustainability targets,” said Costello. “We’re helping our customers with leadership in intelligent infrastructure, powered by our seventh-generation TPUs and geospatial insights via Google Earth Engine.”
As AI matures from hype to production-grade impact, Google Cloud’s strategy is clear: put tools in developers’ hands, embed AI into everyday business workflows, and build partnerships that prioritise trust, scalability and innovation.
“The UK tech startup ecosystem is absolutely booming,” said Costello. “Since 2023, over 60% of Gen AI startups here are Google Cloud customers.”
RELATED STORIES
