Racetrac's data science head on AI-driven operations at 700 convenience stores: from gut instinct to predictive analytics

Summary

Matt Jacoby, head of data science and analytics at RaceTrac, laid out how a 95-year-old Southeast convenience retailer with roughly 700 locations, 10,000 employees, and a portfolio spanning RaceTrac, Raceway, and Gulf is systematically replacing instinct-driven operations with predictive analytics via a Palantir implementation.

The core transformation thesis is moving from gut-based and tribal-knowledge decision-making through four analytics layers: descriptive reporting, diagnostic analysis, predictive modeling, and prescriptive action. Jacoby's data science team focuses exclusively on the latter two. The most operationally critical use case today is not inventory but labor allocation, ensuring the right number of employees with the right skill sets are in each store to meet real-time customer demand. Fresh food offerings, including pizza, sandwiches, and breakfast items, make staffing complexity acute.

A second priority is predictive maintenance on fuel infrastructure. The model flags pumps predicted to fail within two to three weeks, enabling proactive repair rather than reactive outage management. From a customer experience standpoint, Jacoby argues a slow-dispensing pump is worse than a bagged-out one because it traps customers without signaling the problem upfront.

Data architecture has been a strategic enabler. RaceTrac's IT team centralized all data into a unified lake, eliminating the point-of-sale and inventory system interoperability problems that plague many multi-platform retail operators. That foundation is what allows Palantir to function at scale across the organization.

The Palantir deployment is currently managerial, with analytics surfacing at the store support center and store manager level. Jacoby sees a clear path toward putting AI tooling directly in the hands of frontline associates, framing that as a near-term competitive differentiator rather than a long-term ambition.

On EV infrastructure, RaceTrac is building its own charging venues rather than partnering out, betting that longer dwell times from EV charging create a more captive in-store customer rather than a threat to the convenience model. The NACS charging standard consolidation is reducing infrastructure risk industry-wide and accelerating that investment thesis.

RaceTrac ranks top five among privately held companies in Georgia and top 15 in the United States by that measure, context that underscores why the operational leverage from predictive analytics compounds quickly across hundreds of locations.