S&P Global AI Chief Eyes ‘Exciting’ Pace of Agentic AI Innovation

S&P Global, the financial information giant, sees agentic artificial intelligence (AI) as the next transformational wave that can raise business analysis and workflows to new heights, according to its Chief AI Officer Bhavesh Dayalji.

“AI is going to touch almost every single part of the industry,” Dayalji said in an exclusive interview with PYMNTS. “We’re only at the beginning.”

S&P Global’s AI journey actually began decades ago. But the acquisition of Kensho Technologies in 2018, where Dayalji was CEO, accelerated the process and laid a data-ready foundation for the generative AI era, which is still in its infancy.

Since then, the data and insights company has rolled out client-friendly AI tools such as ChatIQ, a generative AI assistant that leverages large language models (LLMs) and is trained on company data to perform such things as industry analysis, strategy research and similar tasks.

The company also embedded GenAI into its S&P Global Marketplace so users can find the data that they need. “If you think about all the insights and data that’s trapped in these structured and unstructured documents, now you can surface them in a very natural language query way,” he said.

For employees, S&P Global has introduced Spark Assist, which is similar to ChatGPT and connected to the company’s data. All of the company’s employees can access it to accelerate their work and workflows.

Spark Assist also helps share learnings across the company’s global footprint by giving employees the ability to share these “sparks.”

“There are things that we’re doing in one country that you know folks from another country can benefit from,” Dayalji said. “So we’ve created a way in which you can share things, or sparks.”

The AI chief also wants to explore how to use a tool like Spark Assist to create “network effects” where the benefits ripple outward to more folks.

Another AI initiative for employees was an AI training program — with specific, in-person programs for the most senior executives, who not only learned about the tools but also the implications for their business units.

Read more: Beyond the Hype: What CFOs Should Know About AI Agents

As S&P Global started to create chatbots to find ways to connect their data, the company found out its clients were all doing the same things.

S&P Global created the Kensho LLM-ready API, which lets clients integrate its datasets into their GenAI models. The API is compatible with OpenAI’s GPT series, Google’s Gemini, Anthropic’s Claude and others. For example, some of its commodities data shows up in Microsoft Copilot.

The next phase for S&P Global is “ensuring that our data shows up in these new ecosystems that are being developed,” Dayalji said.

When it comes to deep research, S&P Global recently launched its Kensho grounding agent that lets users write such things as equity analyst reports and other similarly in-depth analysis pieces.

As AI agents work with other AI agents, they can bring about new insights that have not surfaced before. “They can do interactions and kind of communicate with each other. It just allows for more sophisticated work to be done with AI,” the executive said.

The grounding agent also helps prevent hallucinations by the way it’s architected, by grounding it in truth, which is especially critical for highly regulated industries such as financial services.

Dayalji said S&P Global uses commercially available AI models, depending on the use case. The company did experiment with building its own small language model, which they found to be “very efficient” and is being used for some internal tasks. Other projects include a text-to-SQL agent that let users query databases in natural language.

However, when asked if AI will eventually predict the market’s direction, he said, “The world is interconnected in ways we don’t fully understand. And even AI may struggle to do that because you often have black swan events.” However, AI can certainly guide one’s thinking about it.

To be sure, AI is already being used for trading. The Bank of England, the U.K.’s central bank, recently warned that its use for algorithmic trading could exacerbate market volatility and amplify financial instability.

For S&P Global, the shift to agentic AI has support from the highest levels of the organization. “Our CEO Martina Cheung has GenAI at the forefront of her mind,” Dayalji said, adding that “The companies that embrace that shift first are the ones that are going to win.”

On a personal note, Dayalji said he’s excited to be involved in AI at this pivotal point in time. He has been working on AI for many years, but “My parents didn’t know what I did [and] had no interest.” After ChatGPT, “when I have my parents asking me about AI, it’s exciting.”

The post S&P Global AI Chief Eyes ‘Exciting’ Pace of Agentic AI Innovation appeared first on PYMNTS.com.