AI agents are not software with well-defined inputs and deterministic outputs. These features include connecting language models with enterprise data sources and integrating them with workflows. While many organizations use agents embedded in CRM, ERP, and other employee workflow systems, some companies are experimenting with AI customer experiences.
The speed of deployment is concerning many experts. Many experts can quickly lead to waste, create security risks, and lead to an increase in technical debt. In another study, 82% of cloud experts agreed that AI drives cloud complexity and spending, while 45% said they didn't adequately optimize AI-related cloud usage.
“Companies that are particularly rushing to deploy AI agents through public LLMs,” said Raj Balasundaram, global VP of AI Innovation for Customers at Verint. “Public missteps include pushing unnetworked models into production, exposure to third-party platforms for sensitive data, and lack of observability to track performance, equity or compliance.”
