[ad_1]
Nvidia is seeing continued adoption of agentic AI and expects multi-agentic AI workflows to be deployed at a vast scale in 2026.
“For 2026, we are seeing a real migration from the use of agentic AI, from kind of single agent tasks to orchestrated multi-agent activities,” Kevin Levitt, global business development lead for financial services at Nvidia, told FinAi News.
A multi-agentic AI workflow is just what it sounds like, Levitt said: Multiple agentic AIs working together to complete a task rather than one agentic AI doing a long process.
A multi-agent workflow can improve efficiency and accuracy, he added.
Nvidia deployed a multi-agent workflow this year to aid RBC’s wealth management and investment bankers in developing reports and doing market research, Levitt said, adding that the bank has seen:
- Agents processing 10 times more documents;
- 60% faster report generation; and
- The timeline for alpha pattern discovery being shrunk by nearly 80%.
Nvidia deployed 12 agents for RBC where one agent is responsible for summarizing an earnings call while another is responsible for taking the summary of that information to do an alpha pattern discovery to find tradeable signals, and a third looks for relevant information from thousands of documents including past filings, Levitt said.
End-to-end stack
San Francisco-based Nvidia is seeing high demand from financial services clients for its end-to-end services rather than banks trying to acquire chips and building their own tech stack, he said.
“We’re absolutely seeing full-stack adoption within financial services,” Levitt said. By deploying Nvidia’s services end to end, banks can become more efficient and generate better ROI, he said.
“Financial services [companies] don’t build widgets,” Levitt said. “How they differentiate is how they underwrite, how they acquire customers, how they service customers, and all that’s predicated on understanding the data associated with their customers.”
Nvidia’s end-to-end stack can help FIs understand their datasets better, finding insights quicker in a secure environment, he said.
AI factories are a great example of how Nvidia’s end-to-end services are helping banks by marrying their data with Nvidia’s full tech stack, Levitt said.
Growing data center demand
Banks are becoming AI factory-centric because there are thousands of AI use cases, Levitt said.
Nvidia posted a record $51 billion in revenue from its data centers, up 66% year over year, according to its third-quarter earnings report released Nov. 19.
Data center or cloud revenue is on an annualized run rate of $200 billion a year, much higher than other major cloud providers including, Amazon Web Services ($130 billion), Microsoft Azure ($75 billion) and Google Cloud ($61 billion), according to the companies’ earnings reports.
Nvidia is “winning in [the] data center networking” market as most AI deployments now include Nvidia’s networking and chips, Chief Financial Officer Colette Kress said during the company’s earnings call on Nov. 19.
Register here for early-bird pricing for the inaugural FinAi Banking Summit 2026, taking place March 2-3 in Denver. View the full event agenda here.
[ad_2]
Vaidik Trivedi
Source link