Artificial intelligence is entering a new phase. The challenge is no longer simply to develop systems capable of performing tasks autonomously, but to create the conditions needed to trust their outputs.

As Marco Sodero, Head of Business Development at Bludigit, highlighted in his interview with The Future of Utilities, trust in an AI agent is not immediate. It develops over time through the validation of its responses, comparison with real-world data, and the gradual integration of the technology into business processes.

This journey requires deep digitalisation: not merely replacing paper with digital tools, but redesigning processes and ensuring the availability of structured, reliable, and accessible data.

WorkOnSite is a tangible example of this approach. The solution collects photographs, notes, and checklists directly from field teams and, through WoS AI, automatically verifies whether activities comply with project specifications.

This provides greater visibility into field operations, improves quality and safety, and reduces the need for physical inspections. The results also show a significant impact on operational behaviours: during the early stages of adoption, the rate of compliant checks can increase from approximately 55% to over 95% within a short period.

The true value of agentic AI therefore emerges when technology is no longer simply an available tool, but becomes a trusted component of the operating model.

Watch the full interview and discover how to build trust in AI agents