AI and digital transformation for EV charging networks

Tesla built the Supercharger network as a vertically integrated complement to its vehicle business, and through tight hardware-software integration, ownership of the payment and identification stack (plug-and-charge), and concentrated capital deployment, achieved uptime and customer-experience metrics meaningfully better than the public US fast-charging average. By 2024, most major North American OEMs (Ford, GM, Rivian, Volvo, Polestar, Mercedes-Benz, Honda, Hyundai-Kia, Toyota and others) had signed agreements to adopt the NACS connector and access the Supercharger network — effectively making Tesla's network the default fast-charging standard in North America.

ChargePoint is one of the largest EV charging network operators in North America and Europe, primarily through a hardware-and-software model (selling chargers to site hosts and operating the network management software) rather than the asset-ownership model. The company went public via SPAC in 2021 at a high valuation, faced subsequent profitability pressure as the public-charging market matured slower than expected, and has been working through a multi-year operating-cost rationalization. The category lesson is that the hardware-and-software charging model has different economics from the asset-ownership model — lower capex but lower per-kWh capture — and companies need to be honest about which model they are running.

A regional DC fast-charging operator with 180 sites across four states was running blended uptime around 84%, average utilization around 9%, and was about to fail two NEVI grant compliance checks. We rebuilt the predictive-maintenance pipeline using charger telemetry, deployed an energy-management system to mitigate demand charges at the worst-exposed sites, retired six chronically underperforming sites, and rewrote the field-service operating model around mean-time-to-repair targets.