GE Appliances Puts 800 AI Agents to Work in Its Factories and Supply Chain

GE Appliances has deployed more than 800 AI agents across its manufacturing, logistics, and supply chain operations using Google Cloud's Gemini Enterprise platform, the companies announced April 22 at Google Cloud Next '26 in Las Vegas.

That's not a pilot program.

The Louisville-based manufacturer, owned by Haier, says agents now run inside its Brilliant Factory data platform monitoring equipment health in real time, generating shift summaries in minutes instead of hours, and giving plant managers live visibility into line yields. A Supplier Collaboration Agent handles routine communication with more than 600 suppliers. On the service side, the company moves 27 million parts a year and coordinates with more than 700 service parts suppliers, and it credits the AI buildout with a 25% reduction in parts backorders in 2025.

"AI is now integral to the way work gets done at GE Appliances," said Mandar Deo, the company's vice president of digital technology and chief digital officer.

Quality is part of the pitch too. A Quality Insights Assistant mines customer feedback for defect patterns, which Marcia Brey, vice president of logistics, said means "we can catch defects sooner and improve product quality, ultimately delivering a better consumer experience." Matt Renner, Google Cloud's president and chief revenue officer, said GE Appliances is "a model for the agentic enterprise."

Why AI Manufacturing Matters in the Service Van

The backorder number is the one to circle. A 25% cut in backorders, sustained across 27 million parts a year, is the difference between closing a ticket this week and eating a second truck roll. Parts forecasting has always been the weak link between a correct diagnosis and a completed repair. If agentic forecasting holds up at this scale, fill rates on GE and Haier-family parts should keep improving, and that's money in a shop's pocket.

Faster defect detection cuts two ways. When feedback mining catches a failure pattern early, it surfaces as a serial-range service bulletin instead of a slow drip of mystery callbacks. Techs already know the drill of chasing failures that cluster by production date. AI should compress that timeline. The flip side is more running changes mid-cycle, meaning more board revisions and component swaps hiding under the same model number. Documentation has to keep up, and historically it hasn't.

There's also nothing stopping a 10-truck shop from running the same playbook at small scale. The models GE uses on its lines are cousins of the AI predictive maintenance tools showing up in service software, and connected-appliance diagnostics are already changing service calls. An assistant that drafts estimates, summarizes call notes, and flags repeat-failure customers is the garage-band version of what Louisville built.

The announcement lands on top of the company's $3 billion US manufacturing investment, which is doubling heat pump water heater capacity in South Carolina and reshoring range production to Georgia. New plants, staffed in part by software.

Expect the rest of the industry to follow, and expect the next AHR Expo cycle to be thick with "agentic" branding. The machines coming off these lines are the ones techs will be servicing for the next 15 years, and they're increasingly built, tracked, and quality-checked by software that never sleeps.

Sources

GE Appliances. (2026, April 22). "GE Appliances Reinvents Manufacturing Operations at Scale with Google Cloud's Gemini Enterprise." GE Appliances Pressroom. pressroom.geappliances.com

Google Cloud. (2026, April 22). "GE Appliances Reinvents Manufacturing Operations at Scale with Google Cloud's Gemini Enterprise." Google Cloud Press Corner. googlecloudpresscorner.com