Enterprise AI
June 4, 2025

Decoding Agentic AI for Enterprises With S&P’s Chief AI Officer

Bhavesh Dayalji, Chief AI Officer at S&P Global and CEO of Kensho, shares how he’s redefining enterprise AI with trusted data and agentic workflows.

This is a summary of an episode of Pioneers, an educational podcast on AI led by our founder. Join 3,000+ business leaders and AI enthusiasts and be the first to know when new episodes go live. Subscribe to our newsletter here.

TL;DR:

  • S&P Global is embedding AI into enterprise workflows using trusted, structured data to mitigate hallucinations and enable reliable actions.
  • AI agents are reshaping knowledge work by automating data retrieval, modeling, and report drafting, freeing humans for higher-value thinking.
  • The Kensho team developed Spark Assist, a generative AI tool used by 40,000 employees to personalize and accelerate internal workflows.
  • S&P Global’s LLM-ready APIs enable traceable, citation-backed data access, laying a foundation for scalable and compliant AI agents.
  • The future is multi-agent collaboration, Bhavesh envisions ecosystems where internal and third-party AI agents work together across enterprises.

Before we dive into the key takeaways from this episode, be sure to catch the full episode here:

Meet Bhavesh - Chief AI Officer at S&P Global and CEO of Kensho

Bhavesh Dayalji is the Chief AI Officer at S&P Global and CEO of Kensho, where he’s leading the charge in integrating advanced AI systems into mission-critical enterprise workflows.

Starting his career at CERN as a software engineer searching for the Higgs boson, Bhavesh has traversed a diverse path, from management consulting to scaling Kensho from a fledgling AI startup to its acquisition by S&P Global in 2018.

At S&P Global, Bhavesh spearheads AI strategy, including the development of Spark Assist, a generative AI tool used by 40,000 employees, and Kensho's LLM-ready APIs, enabling traceable, auditable data workflows.

An advocate for responsible, action-oriented AI, Bhavesh is building a framework for trusted enterprise agents, where internal and external AIs collaborate safely and transparently to drive outcomes at scale.

From CERN to Chief AI Officer: Bhavesh’s Unconventional Path to Leading Enterprise AI

Bhavesh Dayalji’s journey to becoming Chief AI Officer at S&P Global and CEO of Kensho is anything but traditional.

Starting as a software engineer at CERN searching for the Higgs boson, he transitioned into management consulting before joining Kensho in its early startup days.

“Like most people's career journeys... I never expected to be doing this,” Bhavesh explains.

At Kensho, he wore many hats across product, engineering, and customer engagement, eventually helping lead the company through its acquisition by S&P Global in 2018.

That move positioned him to shape AI strategy across a 160-year-old enterprise.

“It was really exciting to realize the importance that data has to play in AI, and increasingly the importance AI has to play in S&P Global’s future,” says Bhavesh.

AI Agents in Action: Moving Beyond Chatbots to Automated Enterprise Workflows

While the world fixated on chatbots and copilots after the ChatGPT boom, Bhavesh Dayalji is focused on the next leap: AI agents that can take meaningful actions.

“We think of agents as advanced AI systems that leverage large language models to navigate and complete complex workflows,” he says.

Unlike prompt-based tools, these systems automate tasks like drafting financial reports, retrieving structured data, or building equity analyst models.

Bhavesh notes, “The work that AI agents are taking away is structuring that data, retrieving that data, maybe contextualizing that data.” By doing this groundwork, agents empower humans to focus on higher-order thinking.

“The thing that I was most fascinated with is… that's the reason we work long hours—it isn't because of complex work, it's because of collecting and structuring information.” — Bhavesh Dayalji

He sees a future where everyone has “a small army of specialized AI agents” working behind the scenes to accelerate decisions and value creation.

Trust, Traceability, and the Foundation for Reliable AI

In regulated industries, trust is everything. Bhavesh emphasizes that data integrity and transparency must come before automation.

“You don’t want the AI to take improper actions and then that impacts downstream work,” he explains. That’s why Kensho built grounding tools like the LLM-ready API and grounding agents that pull from structured, auditable datasets.

“It’s not a hallucination. It’s not something that’s made up,” says Bhavesh. Each data point is sourced, cited, and traceable.

This is essential not only for enterprise adoption, but also for building confidence in outputs. “If the trust isn’t there, they’re not really going to run with the outputs,” he adds.

Building this foundation enables safer automation and establishes S&P Global as a trusted partner in the evolving AI landscape.

Scaling Innovation Internally Before Taking It to Market

Rather than launching products prematurely, Bhavesh’s team incubates tools internally first. Kensho’s Spark Assist, a generative AI platform, is now used by 40,000 S&P Global employees to enhance productivity.

“We created our own generative AI tool... so that our people have access to the latest model developments,” says Bhavesh. It lets teams experiment with report generation, research, and analysis using real-time AI.

“If they can’t touch it, if they can’t use it... what use is it?” he asks. These experiments often reveal unexpected value.

“Customers have been blown away by the sophistication of our tool,” he shares. That feedback loop drives external offerings, echoing Amazon’s path from internal tools to AWS.

“It’s about democratizing this technology,” Bhavesh says, “so that it can really impact all of our people and our customers.”

The Future of Work: AI Agents Collaborating Across Teams and Companies

Bhavesh sees the future not just as AI assisting humans, but as agents working with other agents. “We are developing an enterprise agentic framework,” he reveals.

This framework defines how agents interact internally and externally, creating the foundation for multi-agent ecosystems. “It starts to become so easy in which you can start to do end-to-end workflows,” he explains.

“’It’s very clear the moment we have this kind of framework, it starts to become easy to do end-to-end workflows.” — Bhavesh Dayalji

He compares this shift to the early days of the internet, where standards like HTTP unlocked global communication.

Bhavesh predicts a future where S&P Global agents collaborate with customer-owned or third-party agents in secure, structured ways.

“There could be agents we build and others we buy. Is it by the task? Is it by the hour?” he asks.

Whatever the model, the impact will be transformative for how organizations operate.

Would you like to learn more about data accessibility and how it drives success for enterprises? Check out this episode on data readiness & AI with David Aaronson.

Book a 30-minute demo

Explore how our agentic AI can automate your workflows and boost profitability.

Get answers to all your questions

Discuss pricing & project roadmap

See how AI Agents work in real time

Learn AgentFlow manages all your agentic workflows

Uncover the best AI use cases for your business