S&P Global Energy Reduces Conversational Data Product Development Time to Days

S&P Global Energy has successfully reduced the time-to-market for conversational data products from months to days by utilizing Databricks Genie Agents and the Model Context Protocol (MCP). This transformation allows subject matter experts (SMEs) to create and curate Genie Agents corresponding to new dataset groups or commodities, including Chemicals, Crude Oil, Refined Products, Gas & Power, and Liquified Natural Gas (LNG).

The LNG dataset encompasses various components such as facility specifications, cargos, outages, supply and demand fundamentals, netbacks, historical and forecast prices, and contracts. Similarly, the Chemicals dataset includes capacity, production, utilization, trade, demand by end use and derivative, inventory change, and country- and region-level supply–demand balances.

Priyanka John, Vice President of S&P Global Energy, stated, "Genie Agents let our domain experts productize their knowledge of the data directly," highlighting the shift from a traditional engineering backlog to a model where SMEs can publish access to data autonomously. This approach ensures that every answer generated by the agents remains within the company's governance boundaries, as John emphasized, "Making data available to AI did not mean making it available outside our controls."

The new system allows for measurable answer quality, with the transition to Genie Agents significantly streamlining the development cycle. John noted, "What used to take a full development cycle now takes days, and every answer stays inside our governance boundary." This development follows S&P Global Energy's ongoing efforts to enhance data accessibility while maintaining strict control, aligning with previous advancements in agent-based systems reported by Databricks.

As a result, practitioners can expect faster deployment of conversational data solutions, enabling more agile responses to market changes and data inquiries.