NiCE: How AI Agents Boosts Customer Engagement at Openreach
Agentic AI is revolutionising how telecoms operate and the latest in line is Openreach – a unit of the British Telecom Group.
Openreach is reshaping how it delivers broadband upgrades across the UK by using AI agents to drive a more proactive and transparent customer experience.
The company’s partner in this journey – one of the largest digital infrastructure transformation programmes in the UK – is NiCE.
Openreach has deployed proactive AI agents from NiCE Cognigy to redesign customer engagement across 15 million customer journeys.
In this major shift from traditional, reactive service models, Openreach is now anticipating customer needs and initiating conversations in real time.
“Our proactive AI agents are transforming the broadband experience across the UK and delivering tens of millions in financial benefits for Openreach and our customers,” says Chris Herbert, Openreach’s Director of Customer Service.
“By moving to proactive, AI-driven engagement, we’ve improved appointment success, optimised operations and given customers greater clarity during a complex national upgrade.
“At this scale, that combination of efficiency and trust makes a meaningful difference.”
This shift has already delivered measurable results, including a one-third reduction in missed appointments and inbound contact volumes, while improving operational efficiency.
AI-powered engagement model
At the core of this transformation is an AI-powered engagement model that removes friction from the broadband upgrade process.
Using live performance signals, the system adapts in real time to deliver timely updates, suggest next steps and resolve issues before they escalate.
In the model, customers are immediately contacted through their preferred channels – including text message, email and voice – thereby delivering updates before they reach out.
This proactive approach not only improves convenience but also builds trust by making the process more transparent and predictable.
The impact is particularly significant given the scale of Openreach’s operations. Managing millions of customer interactions during a nationwide upgrade would traditionally place enormous strain on service teams.
By automating routine communication and actions, agentic AI enables Openreach to maintain consistency while reducing pressure on human agents.
Boosting operational efficiency
The benefits of proactive AI extend beyond customer experience into core operations.
By cutting repeat contacts and simplifying interactions, Openreach has freed up service teams to focus on more complex and high-value tasks.
Inbound contact volumes have dropped significantly, easing demand on support channels while improving response times where human intervention is required.
Consequently, appointment success rates have increased thereby helping to streamline field operations and reduce costly rescheduling.
“AI is redefining how organisations engage customers, shifting from reactive service to intelligent, proactive orchestration,” says Jeff Comstock, President of CX Product and Technology at NiCE.
“This deployment demonstrates how agentic AI can deliver measurable outcomes at scale, enabling organisations to automate complex interactions while maintaining trust, inclusivity and control.”
AI drives measurable results
The shift to proactive AI engagement is also delivering a clear uplift in customer satisfaction.
Openreach reports a significant improvement in its Trustpilot rating, which has risen to 4.7 out of 5 from its initial 2.0 – all based on hundreds of thousands of reviews following the rollout.
The improvement by NiCE Cognigy’s agentic AI reflects a broader change in how customers experience broadband upgrades.
Faster communication, clearer expectations and fewer disruptions are helping to turn what was once a frustrating process into a more seamless journey.
As organisations across the telecoms sector look to modernise their operations, Openreach’s approach highlights the growing importance of AI in delivering scalable, customer-focused outcomes.


