Google Cloud is expanding its financial services specific agentic AI tools as it makes a deeper push into the industry to drive value for its clients — and it looked to a traditional bank for expertise.   

As part of its Gemini Enterprise for Financial Services program, Google Cloud launched its Financial Research Agent, which can perform tasks including: 

• Know Your Customer standards;

  

• Market analysis;  

• Finding credit opportunities in existing portfolios; 

• Portfolio optimization; and  

• Accelerated bond issuance. 

Google Cloud leaned on Deutsche Bank to understand the workflows and develop tools for maximum impact, Georgina Bulkeley, global head of financial services at Google Cloud, told FinAi News, adding that the agent comes with 50 purpose-built tools for easy deployment. 

“We’re seeing efficiency proof points in client research and bond risk analysis,” Bulkeley said. “Bond-portfolio risk analysis can be reduced to a sub-five-minute workflow, including automated duration hedging suggestions, while research and client-pitch preparation can move from days to minutes.” 

One of the biggest hurdles for AI in financial services is compliance,  Bulkeley said. 

The Financial Research Agent provides transparent methodologies, confidence scores, audit-ready data snapshots and precise citations back to verifiable enterprise data, she said. 

“That’s precisely why design partners like Deutsche Bank worked alongside us — to ensure the platform satisfied rigorous data protection, data residency and governance requirements so financial institutions can deploy this safely and at scale,” Bulkeley said. 

Deutsche as a design partner 

One of the first uses of the agent is within the corporate banking business at Deutsche, a Deutsche spokesperson told FinAi News.  

The $1 trillion bank is using the agentic tool to monitor and analyze mid-sized companies in Germany to uncover credit or other opportunities.  

The workflows in which the agent is deployed include internal briefing memoranda, financial analysis, wallet analysis and client opportunity identification aimed at supporting better research and preparation. 

While developing the agentic platform, the aim was to support internal research and preparation, not to replace human judgement or make autonomous client decisions, the spokesperson said. 

The Financial Research Agent is designed to support regulated workflows by producing structured, traceable and reviewable outputs, the spokesperson said.  

Compared with more generic AI tools, the focus is on consistency, verifiable sources, access controls, audit logging, encryption and data residency, the spokesperson added. 

While the bank’s immediate focus is the initial corporate bank use case, there is opportunity to deploy the underlying tech in areas such as investment banking and private banking, as well as financial crime risk management, advanced forecasting and scenario analysis. 

“Our differentiation is in the combination of out-of-the-box financial specialization and governed execution,” Bulkeley said. 

The bank is slowly rolling out the agent and aims for mass deployment in the fourth quarter, the spokesperson said. Deutsche Bank declined to quantify efficiency gained or costs saved in its early limited roll out of the tool. 

Register here for the FinAi Lending Summit, set for Oct. 7-8 in Las Vegas. This inaugural event will include speakers from Fifth Third and Capital One, as well as a fireside chat with Piermont Bank founder and Chief Executive Wendy Cai-Lee.