Iberdrola deploys Amazon Bedrock AI agents to streamline IT operations

Spanish energy giant Iberdrola is turning to machine logic. The company, which serves millions of customers worldwide, is integrating Amazon Bedrock AI agents directly into its ServiceNow-based IT infrastructure. This is not just another chatbot rollout: it is a full-scale transformation in which artificial intelligence starts handling three critical tasks on its own at the same time. The result: IT workflows speed up, data-processing quality rises, and people get a real chance to focus on strategic issues instead of routine.

Iberdrola's project did not appear out of nowhere. In large energy corporations, IT operations are a huge and complex organism. Thousands of employees submit change requests, incidents arise that require urgent root-cause analysis, and a suitable project management model must be chosen from dozens of options. All of this used to require manual intervention: an analyst looks at a request, tries to understand the context, manually enriches the data with information from various systems, then decides which queue to route it to. Processing speed was limited by the human factor, and quality depended on how attentive an employee was at the end of the working day.

Amazon Bedrock AgentCore let Iberdrola turn this chain into an automated process. The first area is validation of change requests. When an IT specialist creates a draft request, the agent analyzes it instantly: it checks that the data is complete, the description is correct and the request meets corporate standards. There is no need to wait for a person to remember all the mandatory fields and the right format. The agent does this in parallel with data entry and points out what was missed.

The second area is incident management. When a system failure occurs, IT staff hurriedly open a ticket, but the context often stays vague. Have similar problems occurred before? Which systems are involved? Are there known workarounds? Previously, an analyst wandered through various knowledge bases and logs. Now the agent enriches the incident with relevant information on its own: it pulls historical data, links it to similar cases and suggests possible solutions. This shortens the time to first response and raises the likelihood of a correct resolution on the first attempt.

The third component addresses the problem of choosing among numerous management methodologies. In a large organization, teams using Waterfall, Agile, Kanban or hybrid approaches may work at the same time. Which model should a new project use? A conversational AI talks with the initiator, asks clarifying questions about the nature of the work, deadlines and risks, then automatically recommends the optimal approach. This eliminates lengthy meetings and bureaucratic procedures.

For Iberdrola, the effect shows up as shorter ticket-closure times and more consistent data across the organization. Every ticket is now handled to a single standard, without omissions or errors. But it is also a signal for the industry: the era of simple chatbots answering standard questions is ending. Complex agentic systems that can autonomously perform multistep tasks in a complex corporate environment are replacing them. Amazon Bedrock opens the way for such systems, letting companies plug in powerful language models without a data center of their own.

Scaling is the question ahead. Will this approach work in other Iberdrola divisions? Will other energy companies and, more broadly, industrial corporations be able to apply similar solutions? Judging by how quickly interest in Amazon Bedrock is growing, the world of corporate IT operations is on the threshold of a deep transformation.