Day Two - 14 October 2026

10:00 am - 10:30 am EST Masterclass | From prompting to context engineering: Upskilling your team in the age of AI agents

Although AI models have become better at interpreting natural language, prompt engineering is still a critical part of using AI successfully in business. It has shifted away from single instructions towards controlling everything the model can see and act on. In this, prompt engineering has been absorbed into something larger.

CX Network research found that 36 percent of practitioners believe upskilling requirements as AI's biggest impact on their organizations, and our Advisory Board has repeatedly told us that this is the practical help teams are looking for. 

In this masterclass session, we'll focus on what CX teams need to be good at. We'll cover knowledge base architecture, because retrieval failure is often what causes agent errors, and because weak help pages exacerbate customer frustration. We'll look at how to build an evaluation system from your own data, which is where contact center teams have an advantage since they know customers best. We'll also cover how to identify low confidence in the organization and how to address it. Finally, we'll explore the issue of judging output quality. You'll leave with prompt patterns that can be applied straightaway. 

Attendees will learn:
  • The skills that determine whether a CX agent deployment will succeed and how to build them in your team
  • How to construct and maintain an evaluation set from your own contact data, including edge cases and difficult contacts
  • Practical prompt and context approaches for customer use that you can takeaway and use immediately 

10:30 am - 11:00 am EST Getting legal & risk on-board with agentic AI in CX

Since August 2026, Article 50 of the EU AI Act has been enforceable. This means that if your business that if EU customers can reach your customer-facing AI, disclosure is now a legal requirement. There's been a lot of confusion about this because of the Digital Omnibus, which came into force in July 2026, deferring the high-risk obligations element to December 2027, generating a wave of coverage about "delays" to the EU AI Act. But the reality is that Article 50 is now here and the penalties for non-compliance can be significant. 


In this session, we'll explain what CX teams need to get an AI deployment past legal sign-off. We'll set out what Article 50 requires in practice and work through the terms that matter but are often missed. Deloitte found that, while 74 percent of companies plan to deploy agentic AI in the next 2 years, yet only 21 percent have mature governance models to support such an implementation. In this session, we'll talk about how to close that gap. 

Attendees will learn:
  • What Article 50 actually means for customer facing AI in practice and how the Digital Omnibus affects this
  • The questions that surface the terms to look out for during agentic AI discovery phases
  • How to approach governance from a CX standpoint to mitigate legal and risk objections

11:00 am - 11:30 am EST What does agentic AI really mean for the future of CX - and your team?

Employment of customer service reps in the US has declined almost 5 percent for two years in a row. This is a measurable decline but nowhere near the "customer service agents will disappear entirely" narrative that has been regularly touted. Forrester's view is that AI is disrupting the old rule of more contacts = more people needed to handle them, which is changing how teams are shaped. Gartner goes even further, predicting that half of the companies that have cut roles in the name of AI will be rehiring by 2027. Either way, what's clear is that organizations are rearranging the workforce.


In this session, we'll look at what this actually means for the way CX and contact centre teams are set up. We'll talk about which work AI agents are genuinely taking on and why the interactions left for human workers are getting both more difficult and more valuable. We'll also explore shifting roles and responsibilities under AI, including for training and improving AI, knowledge base maintenance and quality assurance. And, critically, we wont dodge the most uncomfortable question: if entry-level role are where CX and service leaders have always started their careers, what will happen to the next generation when these roles are reduced?

Attendees will learn:

11:30 am - 12:00 pm EST Should I build or buy? Unpicking the debate

When we asked our network how they acquired new AI capabilities in 2025, 74 percent said they worked with third-party vendors. In 2026, that figure has fallen to 50 percent. That is a significant shift in a single year and reflects the complexity in the AI vendor market. Functionality across vendors has converged and differentiation is largely about pricing and integration capabilities. Consolidation, too, is increasing, with several high-profile acquisitions in 2026. In this market, buyers are right to be more cautious. Building, however, is not the straightforward alternative. Once a concept has been built, most organizations simply don't have the capacity or internal expertise to scale or maintain the systems. 


In this session we'll unpick the build vs buy decision honestly, with input from practitioners on both sides. We'll look at the variables that actually determine the decision, such as contact volume, target market nuances, ongoing maintenance and long-term scalability. We'll also challenge the binary debate. Many businesses take a blended approach, and Deloitte found that 85 percent of companies expect to customize agents to fit their business, and this capability is increasingly a differentiating factor for vendors. Finally, we'll take a look at what makes a vendor partnership successful even when roadmaps and ownership changes. 

Attendees will learn:
  • The variables that actually drive the decision making in the build vs buy debate and how to identify the most useful ones for your business.
  • What "buying" involves when customization and integration are considered
  • How to maintain a successful and durable vendor partnership 

12:00 pm - 12:30 pm EST AI sovereignty in agentic CX: Where does the data physically live?

CX leaders are increasingly being asked by legal, security and boards where customer data is stored and processed and under which jurisdiction. Most, however, cannot confidently answer that question when it comes to the agentic AI vendors they are evaluating. Deloitte's 2026 State of AI in the Enterprise report found that 77 percent of companies now factor an AI solution's country of origin into vendor selection and that 58 percent now use local vendors. Especially for UKI and European CX teams, choosing between larger, US-headquartered providers is more complex than ever, and smaller, locally-based vendors are gaining an advantage.


In this session we'll explore data residency (where it's physically located), localization (the legal requirement to keep it in a particular place) and sovereignty (which countries laws govern it). We'll also address what makes this an agentic issue: the point is now that an AI agent sends each conversation somewhere to be processed, and that place may be in another country, so knowing where your database sits no longer covers you. 

Attendees will learn:
  • The difference between data residency, localization and sovereignty and why solving one doesn't necessarily solve the others
  • How to find out where your customer conversations are sent once an AI agent handles them and why it matters
  • Procurement questions to ask AI vendors so you get the answers legal, security and the board are asking for