What challenges did AES face before AI implementation?

Increase accuracy and dramatically speed up health and safety audit document reviews. Reduce auditor effort while ensuring reliable, instruction-following agent behavior.

What was the impact of the AI solution?

Reduced audit costs by 99%. Delivered results 99.7% faster, from 14 days to 1 hour. Improved audit accuracy by 10-20% after more than 50 agent audits.

Source

GCPProductionEvidence: Medium65/100

AES deploys genAI agents on Gemini Enterprise Agent Platform to automate safety and compliance audits

Use case typeAudit automationUpdated May 14, 2026

AES deployed gen AI agents to significantly increase the accuracy and speed of health and safety audits. The system processes multilingual audit documents, builds knowledge bases, and coordinates agent tasks with human review in the loop.

Organization: AES
Industry: Energy & Utilities
Location: United States
Published: May 2026

Reported outcomes

−99%

audit cost reductionCost savings

Strategic outcomes

Risk & complianceImproved audit accuracyCost efficiencyAccelerated multilingual audit processing

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Audit cost reduction: 99% decrease

Google Cloud Customer StoryMay 14, 2026Customer storyExplicit claimMedium evidence strength

After conducting more than 50 agent audits, AES saw audit costs reduced by 99%

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: AES
Provider: GCP
Maturity: Production
Linked source: Google Cloud Customer Story

AES deployed gen AI agents to significantly increase the accuracy and speed of health and safety audits

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 2 of 2

1Audit automation
2Document review

Increase accuracy and dramatically speed up health and safety audit document reviews.
Reduce auditor effort while ensuring reliable, instruction-following agent behavior.

Built gen AI agents on Gemini Enterprise Agent Platform using Anthropic Claude models.
Used API calls within the Agent Platform environment to synchronize agents for complex safety audit tasks.
Built knowledge bases, fine-tuned AI agents, and retained human-in-the-loop review.

Technologies

Gemini Enterprise Agent Platform Anthropic Claude 3

Reduced audit costs by 99%.
Delivered results 99.7% faster, from 14 days to 1 hour.
Improved audit accuracy by 10-20% after more than 50 agent audits.

Architecture

AES built gen AI agents on Gemini Enterprise Agent Platform with Anthropic Claude models. The agents process multilingual audit documents, coordinate complex audit tasks, rely on knowledge bases, and keep a human in the loop for review.

Sources & evidence1

Evidence: Medium65/100Evidence strength