Tech giant Google Cloud and chipmaker Nvidia are teaming up to help businesses adapt and deploy agentic AI. 

Google’s AI model Gemini will be deployed on Nvidia’s Blackwell servers at businesses using them, which will enhance the security of sensitive data, including financial transactions and health records, according to an Nvidia’s April 10 release.

“By bringing our Gemini models on premises with NVIDIA Blackwell’s breakthrough performance and confidential computing capabilities, we’re enabling enterprises to unlock the full potential of agentic AI,” Sachin Gupta, vice president and general manager of infrastructure and solutions at Google Cloud, said in the release. 

The Blackwell GPU servers allow businesses to run multimodal AI on site to optimize security, the release stated. 

KPMG, Google Cloud expand AI alliance 

Google Cloud and consultancy KPMG are joining forces to allow KPMG and its clients in financial services, the legal industry and other sectors to deploy scalable AI solutions to improve efficiency. 

Todd Lohr, head of ecosystems and national operations leader for advisory markets at KPMG, told Bank Automation News that gen AI solutions being developed include: 

• KPMG Commercial Lending Assistant: A tool that deploys AI informed with KPMG’s intellectual property and industry knowledge to integrate large-scale commercial lending and payment platforms. 

• Intelligent forecasting: AI-driven automation that enhances financial operations, optimizes procurement decisions and mitigates risk. 

• Risk management solutions: Gen AI-powered internal audit intelligence that accelerates risk identification, anomaly detection and regulatory compliance analysis. 

• AI agents and integration frameworks: Enhance business functions using Google Cloud’s Vertex AI Agent Builder and Agent Development Kit. 

KPMG will use Google’s Agentspace platform to deploy Google Gemini’s capabilities within the organization for efficiency and to develop new products, Rohit Bhat, managing director of financial services and capital markets at Google Cloud, told BAN. Agentspace helps companies develop AI models using a base model like Gemini, Bhat said. 

“In banking, financial services clients will have access to a suite of services and agents built on Google Cloud, providing a single destination to manage their organization’s AI tools and agents,” Lohr said. “This will enable secure and fast data interaction by leveraging KPMG’s Modern Data Platform and pre-built connectors to other enterprise systems.” 

With Agentspace, KPMG plans to streamline access to tools and information, enabling teams to focus on clients, Lohr said. Agentspace will also automate workflows in shared services like procurement, finance and HR, lowering costs and providing differentiated value offerings to clients, he added. 

UWM selects Google Cloud for agentic AI deployment 

United Wholesale Mortgage and Google Cloud created a strategic partnership to modernize mortgage lending processes through AI. 

UWM is using the tech provider’s Vertex AI platform for document extraction and information synthesis to aid underwriters, Jason Bressler, chief technology officer at UWM, told BAN. 

“Our underwriters went from six loans a day to being able to underwrite 14 to 15 a day,” Bressler said. “By summer, we’ll probably be somewhere around 20 loans a day, and by the end of the year, we should be at 30.” 

The platform was customized for UWM, Bressler said, adding that it is 99% accurate. 

Samsung adds Google’s Gemini to home robot 

Smartphone and electronics giant Samsung has selected Gemini AI models to add to its home robot, Ballie, according to a release from Samsung on April 9. 

Ballie can talk to customers and will be able to adjust lighting, greet people at the door, personalize schedules and set reminders, among other things. 

“By pairing Gemini’s powerful multimodal reasoning with Samsung’s AI capabilities in Ballie, we’re leveraging the power of open collaboration to unlock a new era of personalized AI companion — one that moves with users, anticipates their needs, and interacts in more dynamic and meaningful ways than ever before,” Yongjae Kim, executive vice president of the visual display business at Samsung Electronics, said in the release. 

Many tech giants will look to embed other companies’ AI models in their operations like a “pass-through” model, where customer queries will pass through their platform to the base model, Rajul Rana, CTO at IT think tank and consultancy Orion Innovation, told BAN. 

Building a base AI model “is like a $100 million exercise” with continuous upkeep investment, Rana said.  

“Some people are smart enough [not to build their own AI model], because Apple has realized that AI models are becoming a commodity,” Rana said.