_w=800_h=1382.png?v=20230522122229)
A series of Q&As with selected speakers ahead of their appearance at Travolution Connects, the new event from Eventiz by Travelsoft
Sarah Ditton is director of customer success at Alyza, an AI-powered platform that analyses calls, emails, and chats to improve service quality, operational efficiency, and business outcomes. In this Q&A she talks agentic AI, moats and ROI.
Where are we with AI in 2026?
2026 feels like the year where AI has to prove real business value. The industry has moved beyond experimentation, and CEOs and shareholders increasingly expect measurable ROI, operational improvements and clear efficiency gains - not just innovation narratives.
One of the biggest learnings over the last years has been that companies need to start with transparency and a clear vision instead of trying to automate everything immediately. The organizations seeing the most success typically start small, understand their operational realities first and then scale AI step by step. From an Alyza perspective, that shift is very positive because it aligns strongly with our view that understanding customer interactions and operational processes comes before large-scale automation.
Is there anything that has surprised you over the past couple of years, any use cases which you think might have been overlooked or where AI has not added enough value to justify the investment?
One thing that surprised us was how often companies tried to automate processes they did not fully understand. In many cases, AI simply accelerated inefficient workflows instead of improving them. That is why transparency and operational understanding became much more important than many initially expected.
At the same time, some use cases massively outperformed expectations, especially where AI was used to improve human decision-making instead of replacing it entirely. A good example for us was a river cruise company where we were able to improve conversion rates by 89% within six months by analysing customer interactions, identifying successful conversation patterns and systematically improving service and sales processes.
We also believe some of the most underestimated use cases are still hidden in customer service operations themselves: understanding contact drivers, reducing avoidable demand, improving coaching and identifying revenue opportunities directly from customer conversations.
Is there any one development or announcements that was more significant than the others for Alyza?
For us, the biggest shift was clearly the rise of agentic AI. It fundamentally changed what is practically possible and allowed us to build the first version of our AI Assistant in a way that delivers real operational value instead of just acting as another dashboard or chatbot.
It also changed how we think about our own platform architecture. We are currently adapting our backbone and orchestration layers to become both faster and better at connecting insights, recommendations and operational actions across multiple systems and workflows.
For AI-enabled/native businesses (such as Alyza) who have established a foothold in the market, how do you defend your position? What is your moat? How can you maintain differentiation?
Our biggest differentiation is probably not the AI model itself, because foundation models will continue to evolve and commoditize over time. The real moat comes from solving real operational problems together with clients and building proven expertise in complex customer environments.
Our customers expect measurable performance improvements, and over the years we have built a strong track record with real-world results and operational learnings. Only by working deeply with real clients and their day-to-day challenges do you gain the understanding needed to build solutions that actually work at scale and create sustainable value.
How does Alyza keep on top of developments within the AI industry itself and how do you assess whether these developments can add value to Alyza or its clients?
We usually do not start from the technology itself, we start from the client problem. The trigger is normally an operational need or a customer challenge, not necessarily the latest AI trend.
A good example is our AI Assistant. Clients had been asking us for years for more proactive and actionable support, but only recent agentic AI approaches made that practically usable in a scalable and reliable way. That is one of our biggest advantages: we stay very close to our clients, understand their operational realities and then evaluate which new technologies are mature enough to solve those problems meaningfully.
Do you have concerns or contingency plans for if and when the AI bubble bursts? What might that bursting bubble look like? Are we too far down the road?
A market correction is definitely possible, especially for projects that are driven more by hype than measurable business impact. We already see companies becoming much more focused on ROI, operational value and sustainable integration into existing business processes.
For us, the situation is slightly different because we are deeply integrated into our clients’ operations and orchestrate different AI technologies to solve concrete business challenges. In many ways, we see ourselves less as a single AI product and more as an operational intelligence and optimization layer built on top of multiple AI capabilities.
Looking ahead, what are you seeing that excites you, for the travel industry and for Alyza? Do you have any idea what the next big thing might be?
What excites me most is the evolution of AI in customer service from pure analytics toward operational improvement and, where appropriate, controlled automation. We are currently bridging the gap between generating insights, improving operational quality and eventually enabling automation on top of trusted knowledge.
One area we find especially exciting is the creation of self-learning knowledge bases from customer interactions. Our goal is to transform millions of real conversations into structured operational knowledge and connect this intelligently with automation systems. A good example are dynamic playbooks for voicebots, that continuously learn from successful human interactions and improve over time in a controlled and measurable way.
Travolution Connects is a new event brand for the UK market from Travolution’s parent company Eventiz by Travelsoft. It is an invitation-only senior leaders event, with the launch event taking place on June 23 in London. For more details, visit the Travolution Connects web site and for sponsorship or guest list queries, please contact andy.hibberd@travolution.com.
Other Travolution Connects Previews:
Roopak Pati, managing director, Oppenheimer & Co Inc
Manuel Hilty, CEO and co-founder, Nezasa
