Southern California Edison × Microsoft — Customer Service Automation

Southern California Edison

Customer Service Automation · Real-time Analytics · Risk Management

Southern California Edison needs to maintain grid reliability as electricity demand rises from electric vehicles, buildings, and data centers, while operating across a large service area where accurate data and safe use of AI are critical.

Reported impact

Business outcome

SCE says it is enhancing grid operations with predictive analytics and AI, helping call center energy advisors today while aiming to use real-time data and forecasting to detect outages earlier, anticipate demand, and prevent disruptions more proactively.

Solution

Southern California Edison described an operational AI program that combines Microsoft Copilot, GPT-based capabilities, predictive analytics, and forecasting models to support both grid operations and customer-facing work. The utility first focuses on data readiness across its 50,000-square-mile footprint so AI and ML models have the right inputs. A new communications infrastructure is being rolled out to bring in real-time operational information from the grid, which then feeds forecasting use cases such as electric-vehicle penetration, grid demand, and broader grid forecasting. On the serving side, SCE is using AI in the call center to help energy advisors and is building a robust knowledge base intended to support a more autonomous conversational chat experience over time. The source does not name a storage platform, but it clearly shows an architecture pattern of real-time ingest, model-driven forecasting, and advisor-facing assistance with governance attention around AI guardrails and information protection.

Data flow

Grid and demand data gathered across SCE's territory, along with real-time information from a new communications infrastructure, is used to feed forecasting and predictive analytics capabilities. The resulting AI outputs support grid forecasting, demand planning, and call-center advisor assistance, with a knowledge base intended to power a more autonomous conversational chat experience.

Solution architecture

2 components · 2 layers

Compute
- GPT Used within SCE's AI capabilities alongside forecasting and call-center support use cases.
Serving
- Microsoft Copilot Used by SCE as part of its current AI tooling to support operational work and experimentation.

Architecture clues

AI use cases mentioned include EV-penetration forecasting, grid forecasting, demand forecasting, and call-center assistance.
SCE is rolling out a new communications infrastructure for real-time grid information.
The source highlights governance concerns such as protecting information and monitoring AI use.
The utility operates across 50,000 square miles.

Evidence from the source

"We all have data challenges—we preside over 50,000 square miles, so accurate data is critical."

"We’re rolling out a new communications infrastructure so we can get real-time information, which will be very useful in managing the increasing demand."

"We’re using Microsoft Copilot, GPT, predictive analytics and forecasting electric vehicle penetration."

"We’ve also put some AI capabilities in grid forecasting and demand as well as in our call center to help our energy advisors."

"With millions of sensors and vast amounts of data we’ll collect—petabytes and zettabytes—we’ll gain the ability to detect outages or even anticipate and prevent them proactively."

Southern California Edison

Business outcome

Solution

Data flow

Solution architecture

Architecture clues

Evidence from the source

How scores are calculated

Sources & provenance

Southern California Edison customer story

Date evidence

Similar use cases