S&P’s intelligence engine operates from Delhi NCR

Delhi NCR is at the heart of S&P Global, the $14-billion data, research and analytics firm. The Manhattan, New York, headquartered company’s key technology operations for thousands of global customers are handled by teams in India. S&P Global has around 15,000 employees in India, more than a third of its global strength, with 5,500 in Gurugram and Noida. The company says India manages more than 300 products and serves clients in financial services, oil & gas, mobility, automobiles, energy and beyond.Frank Tarsillo, CTO of S&P Global Market Intelligence, and his colleague, Dan Bennett, head of technology for S&P Global Enterprise Data Organisation at S&P Global, were in Delhi-NCR recently to oversee delivery operations and expansion activity. That expansion includes a third office in Gurugram to support global clients. “The majority of our product development, especially in market intelligence, happens in Delhi NCR – from premium desktop offerings like CapIQ Pro (a data and analytics platform) to software that supports clients’ workflows,” Tarsillo told us.Teams here handle tasks spanning engineering and data modeling to product design and client support – empowering global industries with data transformed into actionable intelligence.In the energy space, engineers in Delhi NCR build and maintain modeling software used by oil & gas companies for subsurface exploration. Teams here manage global credit ratings, they support automotive supply chains and energy transition analysis. They help global clients analyse sectors and trends, manage risk, and make decisions around maritime, trade, economics and country risk.Tarsillo says the expanded office in Gurugram is to keep people together and foster collaboration. “Gurugram has consistently offered us access to great talent and scale. It’s the cornerstone of our global strategy,” he says.AI central to transformationAI has been a key driver of S&P Global’s transformation. The company’s 40,000 employees use Spark Assist, a proprietary AI assistant whose core engineering team, including its head of engineering, is based out of India. Initially built with GPT models, Spark Assist now supports multiple models like Claude and DeepSeek and has become a go-to tool across functions — from summarising meetings to generating code suggestions.“The uptake (of Spark Assist) was organic,” says Tarsillo. “People started using it to get answers faster, summarise documents, and even route customer service requests intelligently. We saw engineering productivity rise by 10–12% within three months of adoption. More impressively, junior engineers began performing at levels two notches above their experience.”Spark Assist allows employees to create reusable prompt libraries, share workflows and build task-oriented agents. Integrated into tools like Excel and Office 365, it is fast becoming the backbone of S&P’s knowledge-sharing culture. The platform also plays a critical role in document intelligence. For instance, in CapIQ Pro, users can now instantly access summarised insights, sentiment analysis and even suggested questions based on corporate earnings transcripts – without reading through entire documents.“What excites me is not just efficiency but insight,” says Tarsillo. “We’re moving into a space where AI tells you what questions to ask — what you might have missed. That’s game-changing.”But AI can hallucinate, cautions Bennett. “That’s why structured, semantically rich data is crucial. We’re investing in better data governance and semantic layers to ensure AI makes quality decisions,” he says.Going forward, S&P is betting big on AI-driven growth. “More efficiency means more capacity – and that lets us tackle harder problems we couldn’t before,” says Tarsillo.