Improving staff productivity at Enel using Amazon Bedrock

This article discusses how Enel, a leading electric utility company, used Amazon Bedrock and other AWS services to automate ticket management and resolution for IT service desk requests related to business applications.

Specifically, the article covers:

• The context: Enel operates an IT service desk on ServiceNow, and many repetitive non-standard requests for business applications were still manually processed, leading to long resolution times and staff effort that could be optimized.
• The solution: Enel built a retrieval-augmented generation architecture using Amazon Bedrock, Amazon Titan, and Anthropic Claude. This generates appropriate responses for tickets, such as instructions, routing to resolver groups, or closing tickets.
• Business outcomes: The solution reduced case resolution time from 1 day to under 2 minutes, resolved 15% of tickets automatically, and reduced first response time from 9 hours to 1 minute.
• Conclusion: Enel plans to expand the use of foundation models and large language models into more products by integrating Amazon Bedrock deeper into their platform.