VUX World

The two things contact centres have been missing until now

The contact centre holds a record of almost every customer interaction. Every conversation carries a signal about why customers are calling and what is behind it. Processing those conversations at scale took the power the contact centre never had before. Ping Wu, CEO of Cresta, joins us to work through what that shift means for enterprise contact centres, from cutting the calls that should never happen to automating the ones that can be. We get into how AI changed what a contact centre can see. Every call, every screen, every exchange can now be processed rather than sampled. Human effort never scaled the way silicon does, so for years most of that record went unread. Large language models bring processing power and language understanding to work through millions of conversations, see what is driving them and turn that into insight, coaching and automation. Show notes
 Follow Ping on LinkedIn: ⁠https://www.linkedin.com/in/pingwu Learn more about Cresta: https://cresta.com Download Cresta's CX Workforce Report: https://cresta.com/report/cx-workforce Register for our webinar - Meet contact centre 3.0. The playbook you're using is obsolete: vux.ai/3.0 Follow Kane on LinkedIn: ⁠⁠⁠⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠⁠⁠⁠⁠ Find out more about VUX: ⁠https://vux.ai⁠⁠⁠ Subscribe to VUX World: ⁠⁠⁠⁠⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠⁠⁠⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

From chatbot to voice AI: what changes?

Voice AI is the channel everyone is talking about in 2026. One of the biggest things slowing it down is often already sitting inside your stack. In this episode, I’m joined by Endre Davids from Seamly to explore what really happens when businesses take an existing chatbot and move it into the voice channel. We discuss why latency often starts deeper in the conversational platform than teams expect. Endre explains how companies can spend time optimising speech recognition and text-to-speech while the chatbot itself is responsible for much bigger delays. The conversation gets into why voice needs a different interaction model, because people speak differently from how they type. That puts far more pressure on getting the conversation design right. We also cover how to manage perceived latency, when it makes sense to reuse existing chatbot infrastructure and how sentiment, multimodality and customer context could shape richer voice experiences. Endre and I keep landing on the same point. The success of voice AI depends heavily on conversation design, and voice punishes the shortcuts that chat lets you get away with. Show notes
 Follow Endre on LinkedIn: https://www.linkedin.com/in/endredavids Learn more about Seamly: https://seamly.ai Conversational Conference 2026: https://events.frankwatching.com/conversationalconference Follow Kane on LinkedIn: ⁠⁠⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠⁠⁠⁠ Find out more about VUX: https://vux.ai⁠⁠ Subscribe to VUX World: ⁠⁠⁠⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠⁠⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

Cobus Greyling: why intents, graphs and conversation design have come full circle

I'm joined by Cobus Greyling, AI evangelist at Kore.ai, for a wide-ranging conversation on where enterprise AI actually stands today. We explore why many of the ideas now being presented as breakthroughs, from graph engineering and orchestration to intent detection and state management, have deep roots in conversational AI. We also get into one of the biggest barriers to enterprise AI adoption: data. We highlight why customer experience could become the entry point for much wider organisational transformation. We touch on how forward-deployed engineers could help businesses navigate that change and why security will become increasingly important as AI agents gain access to more systems and information. I close with a buzzword bingo round for Cobus covering fleet engineering, agent swarms, orchestration and universal agents, and Cobus explains why he thinks Open Claw was the biggest red herring of the year. Show notes
 Follow Cobus on LinkedIn: https://www.linkedin.com/in/cobusgreyling Cobus's website: https://cobusgreyling.me Follow Cobus on Substack: https://cobusgreyling.substack.com/ Discover more about Kore.ai Article - The Untrainable by Sarah Guo: https://saranormous.substack.com/p/the-untrainable Video - Andrej Karpathy: Software Is Changing (Again): https://www.youtube.com/watch?v=LCEmiRjPEtQ Follow Kane on LinkedIn: ⁠⁠⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠⁠⁠⁠ Find out more about VUX: ⁠⁠https://vux.ai⁠⁠ Subscribe to VUX World: ⁠⁠⁠⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠⁠⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

From answers to actions: building better government AI

Plenty of AI assistants can answer the question in front of them. Fewer can work out what someone is actually trying to do and help them do it. In this episode, we're joined by Pete Kowalczyk, AI Conversation Designer at the Government Digital Service (GDS), to talk about designing AI experiences across the vast and complex landscape of UK government services. Pete explains how GDS is exploring knowledge graphs, persistent state and goal detection to move beyond basic RAG systems. The aim is to connect information with action and guide people through non-linear journeys, such as moving house or registering a birth. One of the key goals is to reduce the burden created by fragmented processes and legacy systems. We discuss why conversational systems need to track user goals, how small interaction loops can be built into rich experiences over time and why transitions between stages of a conversation carry so much weight for trust and usability. We also get into progressive disclosure, deciding when an AI should stay concise and when it should offer more, and how to judge whether a link belongs in a response at all. Pete points to a bank surfacing patterns in someone's spending as an example of how organisations can build proactive features from data they already hold, rather than only responding when asked. Pete shares how the designer role is evolving. Designers now need sufficient technical understanding to prototype and work directly with LLMs through practices like vibe coding, while maintaining the creative ability to understand experiences beyond incremental chatbot improvements. He closes with a warning against the industry's current resolution rut, where success gets defined too narrowly as fast ticket resolution rather than genuinely useful, proactive AI. Show notes
 Follow Pete on LinkedIn: ⁠https://www.linkedin.com/in/petekowalczyk Video of Pete's talk at the Conversational AI Meetup in London: ⁠https://youtu.be/001PeZv3N0k⁠ Lead with Tempo 2026: https://www.leadwithtempo.com/2026schedule London Design Festival: https://londondesignfestival.com Follow Kane on LinkedIn: ⁠⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠⁠⁠ Find out more about VUX: ⁠https://vux.ai⁠ Subscribe to VUX World: ⁠⁠⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

How Simplyhealth scaled voice AI to resolve 30% of calls with Paul Atkins

Almost a third of Simplyhealth's incoming calls are now resolved by AI, and email handling time has fallen from 12 minutes to 2. The company is achieving those gains while maintaining the service standards that placed it 31st in KPMG UK's Customer Experience Excellence report. In this episode, we're joined by Paul Atkins, Head of Business Optimisation at Simplyhealth, the UK's member-owned health cash plan provider serving around 2.5 million customers. Paul explains how Simplyhealth introduced voice AI gradually, tested more than 1,000 real customer calls and built trust across a highly regulated organisation. We also get into how Simplyhealth measures resolution, quality and customer effort, why intensive human listening still matters during rollout, and the organisational call to define what "good enough" looks like rather than letting advisors over-engineer every reply. Paul talks through the role of business readiness teams in driving adoption and why knowledge management could become the next major challenge in making AI work at scale. Show notes
 Follow Paul on LinkedIn: https://www.linkedin.com/in/paul-atkins-1nsight Follow Kane on LinkedIn: ⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠⁠ Find out more about VUX: https://vux.ai Subscribe to VUX World: ⁠⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

Inside the platform enterprises use to build AI agents with Rasmus Hauch, CTO at boost.ai

Safe, reliable, production-ready AI agents don't come from the model. They come from everything built around it. Rasmus Hauch is CTO of Boost.ai, one of the enterprise platforms banks, insurers and public sector organisations use to build customer-facing AI agents. His work sits one layer below every deployment: the orchestration, the guardrails, the testing infrastructure and the model choices that decide whether an agent holds up in front of millions of customers. In this episode, Rasmus opens that layer, explaining why conversational AI is shifting away from traditional conversation flows towards orchestration, testing and continuous evaluation. He walks through how Boost approaches it in practice: persona-based simulation across thousands of digital and voice conversations, predefined test suites for jailbreaking attempts and financial advice, layered guardrails inside and outside the agent, and a trust loop that feeds production results back into the build. Rasmus is direct about the hard parts. We touch on why the big LLM providers can't yet deliver the consistent response times voice demands, which is why Boost runs fine-tuned smaller models on its own infrastructure for latency-critical voice work. Rasmus explains why he advises clients against plugging in their own models. We also discuss where liability sits under the EU AI Act when an agent gets it wrong, and why a volcano eruption in Iceland is a good example of what no model can be trained for. Towards the end, Rasmus shares what's exciting him right now, including adaptive voice, real-time language switching and where agent-to-agent protocols are heading. Show notes
 Discover more about boost.ai Follow Rasmus on LinkedIn: https://www.linkedin.com/in/rasmushauch Follow Kane on LinkedIn: ⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠ Subscribe to VUX World: ⁠⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

How ING is building the instant mortgage with agentic AI

A traditional mortgage can take weeks to process. Every application moves through rounds of document requests, condition checks, and broker queries before a decision is made. ING is using agentic AI to change that. In this episode, Bahadir Yilmaz, Chief Analytics Officer at ING, explains how one of Europe's largest mortgage lenders has implemented agentic AI in its mortgage process. The system is live in the Netherlands, rolling out to Germany. Applications now complete in days, a full day faster since AI was introduced. The goal: the instant mortgage, within the next couple of years. We cover why ING chose its highest-stakes product for its first agentic deployment, why the agent handles operational work while underwriting remains with traditional machine learning, and how the team built a golden dataset of 400 mortgage applications to test a system no one had tested before. Furthermore, Bahadir shares how ING scaled to 114 AI initiatives with 42,000 employees using AI. We discuss why risk management came before ambition, and why only 10 of his 600 experts were ready to build agentic systems. Show notes
 Follow Bahadir on LinkedIn: https://www.linkedin.com/in/bahadir-yilmaz-21170479/ Follow Kane on LinkedIn: ⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠ Take our AI Maturity Assessment: ⁠⁠⁠https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio⁠ Subscribe to VUX World: ⁠⁠⁠https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠https://open.substack.com/pub/kanesimms⁠

From TOBi to Super TOBi. Vodafone's AI Evolution with Jordan Anglin

In this episode, recorded at the CCW UK Summit, we’re joined by Jordan Anglin from Vodafone to explore the company's approach to customer experience and how AI supports faster, more seamless issue resolution across digital and human channels. Jordan explains how Vodafone combines its AI assistant, TOBi, asynchronous messaging, and dedicated resolution teams to reduce customer effort, eliminate repetitive conversations, and create genuine ownership from start to finish. Jordan also shares how AI is improving frontline performance through agent assistance, conversation summaries, quality assurance and real-time guidance, while offering a practical perspective on where large language models deliver real value and where they fall short. Jordan closes with his take on where AI attention in customer experience is currently misplaced, and where he sees low-hanging fruit that's easy to miss. Show notes Discover the CCW Europe Summit in Amsterdam: ⁠https://europe.customercontactweekdigital.com/events-ccweurope/?utm_source=vux.world&utm_medium=Media%20Partner&utm_campaign=27400.010_VUX_Podcast%20episodes%20post%20event&utm_term=&utm_content=&disc=&extTreatId=7638147⁠ Follow Jordan on LinkedIn: ⁠⁠⁠⁠⁠https://www.linkedin.com/in/jordan-anglin-a2672610b/⁠⁠⁠⁠ Follow Kane on LinkedIn: ⁠⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠⁠ Take our AI Maturity Assessment: ⁠⁠⁠⁠https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio⁠⁠ Subscribe to VUX World: ⁠⁠⁠⁠ https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio ⁠⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠⁠ https://open.substack.com/pub/kanesimms⁠

Turning AI hype into real customer value, with Simplyhealth and EdgeTier

We recorded this episode at the CCW UK Summit in May, capturing some of the key conversations shaping customer experience, service and AI. AI is everywhere in customer experience, yet few organisations are turning that excitement into measurable business outcomes. We're joined by Paul Cuglietta, Customer Services Director at Simplyhealth, and Paul Atkins, Head of Business Optimisation at Simplyhealth, alongside Shane Lynn, CEO and Founder of EdgeTier, to explore what successful AI adoption really looks like. We discuss how Simplyhealth scaled generative voice AI with PolyAI from 5% to 50% of interactions and how the company uses EdgeTier's conversational intelligence platform to turn everyday customer conversations into actionable business insights. We also cover why organisations should start with real business problems rather than chasing the latest AI trends. Simplyhealth funnels its conversational data through a prioritisation process to decide which customer journey improvements to make first. The team explains why progress comes from small measurable iterations and how they keep a human in the loop to handle the moments that automation should not touch. We also weigh in on where the AI narrative is overhyped, where its real value remains underutilised, and why the unglamorous work of digging through conversations to fix broken policies and processes is where genuine impact happens. Show notes Discover the CCW Europe Summit in Amsterdam: https://europe.customercontactweekdigital.com/events-ccweurope/?utm_source=vux.world&utm_medium=Media%20Partner&utm_campaign=27400.010_VUX_Podcast%20episodes%20post%20event&utm_term=&utm_content=&disc=&extTreatId=7638147 Follow Kane on LinkedIn: ⁠⁠⁠https://www.linkedin.com/in/kanesimms⁠⁠ Take our AI Maturity Assessment: ⁠⁠⁠https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio⁠ Subscribe to VUX World: ⁠⁠⁠ https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio ⁠ Subscribe to The AI Ultimatum Substack: ⁠⁠⁠ https://open.substack.com/pub/kanesimms⁠

How the company behind Uswitch and Money.co.uk uses AI to resolve up to 70% of customer conversations with Anna Fedorchuk, RVU

When you switch off your call centre, AI has no safety net to fall back on. RVU did exactly that, building an AI-first customer operations function that serves millions of UK consumers. In this episode, we’re joined by Anna Fedorchuk, Customer Operations Manager at RVU, the company behind brands including Uswitch, Confused.com and Money.co.uk. Anna shares how her team transformed customer support from a traditional contact centre into an AI-first operation, achieving AI resolution rates of 65-70% while maintaining a strong customer experience. We discuss how RVU rolled out AI across chat, email and social media without relying on engineering teams, and why building the right knowledge base mattered far more than chasing the latest model. Anna walks us through how AI categorisation and carefully designed workflows lifted resolution rates across genuinely complex customer journeys, where similar keywords across energy, broadband and mobile made accurate routing a real challenge. Anna also describes how Claude and Intercom now work together to analyse conversations and surface operational insights in the team's own language. People are a big part of this conversation. Anna explains how involving her team early gave each agent their own vertical to run and a real sense of ownership, and how AI even softened the stress of handling difficult complaints. She closes on what's next: data connectors and more agentic customer experiences. Show notes
 Follow Anna on LinkedIn: ⁠⁠https://www.linkedin.com/in/anna-fedorchuk-b84232102/⁠ Follow Kane on LinkedIn: ⁠⁠https://www.linkedin.com/in/kanesimms⁠ Take our AI Maturity Assessment: ⁠⁠https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio Subscribe to VUX World: ⁠⁠ https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio Subscribe to The AI Ultimatum Substack: ⁠⁠ https://open.substack.com/pub/kanesimms⁠

Omotenashi: Japanese concept of hospitality and what CX teams can learn from it with Daniel Orenes Ferrández, Uber

Japan has one of the lowest average NPS scores in the world, and the reason is cultural: Japanese customers expect more. Daniel Orenes Ferrández, Senior Manager of Customer Experience at Uber, discusses Omotenashi, the Japanese philosophy of anticipating customer needs and apologising sincerely before resolving a problem. Daniel has lived and worked in Japan for over a decade, including five years leading conversational AI initiatives at Rakuten. He explains why the Japanese language has between ten and fifteen ways of saying sorry, why choosing the right level matters as much as solving the issue, and how generative AI prompts can now select an apology tone based on contact type and severity, something fixed NLU responses could never do. The conversation covers Aimai, the Japanese concept of ambiguity used to maintain conversational harmony. Japanese speakers rarely say no directly, and AI systems need specific prompting to catch these signals in sentiment analysis. We also discuss the escalation strategy, digital mascots, integration with the Japanese messaging app Line, and practical guidance for global brands deploying conversational AI in Japan. Show notes
 Read Daniel's Substack, Japan CX Insider: https://danielferrandez.substack.com/ Follow Daniel on LinkedIn: https://www.linkedin.com/in/danielorenesferrandez/
 Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms
 Take our AI Maturity Assessment: https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio
 Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW?utm_source=podcast&utm_medium=audio
 Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms

The trust layer banks need for an agentic future with Mathias Fanschek, Raiffeisen Bank International

What happens when customers stop opening banking apps and start managing their finances through AI assistants? In this episode, we sit down with Mathias Fanschek, Head of Digital Transformation and Retail Strategy at Raiffeisen Bank International, to explore how one of Europe's largest banking groups is preparing for a future of headless banking, agentic commerce and AI-powered customer experiences. With 18 million customers across 11 markets and managing over €200 billion in assets, Mathias shares what it really takes to move generative AI from experimentation to production. We had a good discussion about virtual mobile assistants, transforming legacy data infrastructure, and managing executive expectations in a rapidly changing, highly regulated landscape. You'll also hear Mathias' perspective on Revolut's new PRAGMA model, why context is the missing ingredient in today's AI experiences and how banks can remain trusted intermediaries as customers increasingly interact through AI-powered distribution channels.

What most companies get wrong about voice AI with Henry Vaage Iversen, boost.ai

We sit down with Henry Vaage Iversen, Co-Founder and CCO at boost.ai, to explore why voice AI is finally reaching mainstream deployment across banks, insurers, telcos and other regulated industries.We dig into what's changed in the technology, why phone volumes continue to rise and how large language models are reshaping what's possible on the voice channel. Henry shares practical lessons from deploying conversational AI at scale, including how some customers are now achieving 60-75% voice automation. We also discuss why the gap between chat and voice experiences still catches companies off guard, and what it really takes to get voice AI right in terms of testing, latency, and conversation design.We explore why layering AI onto legacy channels only gets you so far, and why the real opportunity lies in rethinking customer journeys from scratch. The conversation wraps with a story about what happens when AI agents start calling other AI agents, something our team recently experienced firsthand. Show notes Find out more about boost.ai Follow Henry on LinkedIn Take our updated AI Maturity Assessment Subscribe to our newsletter Subscribe to The AI Ultimatum Substack

• Transcript

How PolyAI built a voice AI model that handles millions of customer calls with Nikola Mrkšić, Co-founder and CEO

Five years after his ⁠first appearance on the podcast⁠, PolyAI CEO Nikola Mrkšić returns to VUX World to discuss how conversational AI has evolved from intent-based systems into enterprise-scale AI agents handling millions of customer interactions. PolyAI began as a research-led spinout from Cambridge University with a singular focus on making automated voice conversations genuinely useful. Since then, the company has grown from a handful of pilot deployments to powering customer service operations for nearly 200 enterprises worldwide, including highly regulated industries such as banking, healthcare, insurance and utilities. In this episode, we explore what it actually takes to build production-ready conversational AI systems at scale. We go deep into PolyAI's proprietary Raven model, the realities of building and training purpose-built language models for enterprise customer service. We also talk about why most businesses underestimate the complexity of deploying AI agents into real-world operations. The conversation also touches on PolyAI's brand campaign starring Gordon Ramsay. Ramsay's restaurants have been a PolyAI client for a few years, and the company launched a campaign timed to the Super Bowl in February 2026 with Ramsay dramatising the contrast between legacy phone menus and PolyAI's voice agents. Nikola explains how the partnership came about and why Ramsay's standards made him a fitting brand ambassador for the company. Show notes Find out more about PolyAI: ⁠https://poly.ai/⁠ Follow Nikola on LinkedIn: ⁠https://www.linkedin.com/in/nikola-mrksic⁠ Follow Kane on LinkedIn: ⁠https://www.linkedin.com/in/kanesimms⁠ Sign up for our upcoming webinar - Cutting headcount isn't an AI strategy: ⁠https://events.vux.ai/getvocal?utm_source=podcast_network&utm_medium=podcast⁠ Take our updated AI Maturity Assessment: ⁠https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast_network&utm_medium=audio⁠ Subscribe to our newsletter: ⁠https://vuxworld.typeform.com/to/Qlo5aaeW/?utm_source=podcast_network&utm_medium=podcast⁠ Subscribe to The AI Ultimatum Substack: ⁠https://open.substack.com/pub/kanesimms

• Transcript

Most AI projects fail. Here's how to pick the ones that don't with Thomas Schaefer, Uniphore

Enterprise AI projects fail more often than they succeed. The MIT figure Thomas Schaefer cites puts that at 95%. The reasons are rarely the model. They are the architecture beneath it, the data that feeds it, and the use case selection that determines whether any of it was worth starting. Thomas Schaefer is Director of Sales Engineering at Uniphore, based in Germany. He has over 20 years in enterprise technology, including 12 years at Jacada across solution architecture, professional services and presales leadership across EMEA and APAC. He joined Uniphore in 2021 and works with enterprise clients on AI transformation, agentic workflow design and the operationalisation of business AI at scale. Show notes Discover more about the CCW UK Summit: https://europe.customercontactweekdig... Find out more about Uniphore: https://www.uniphore.com/ Follow Thomas on LinkedIn: https://www.linkedin.com/in/thomas-schaefer-4027a611/ Follow Kane on LinkedIn: / kanesimms Take our updated AI Maturity Assessment: https://vuxworld.typeform.com/to/a26b... Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5... Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript

Building enterprise AI agents in hours with Merlin Bise, CTO at Inbenta

Enterprise AI tools can spin up a prototype in an hour. Getting that prototype to production with the accuracy, security, and scale a real business needs still takes months for most companies. Inbenta says its new Encore platform closes that gap in days or even hours. Merlin Bise, CTO at Inbenta, returns to VUX World to walk through how Encore builds production-grade AI agents on the fly when a customer deploys a use case. The platform ingests content from websites, documents, recordings and connected systems like SharePoint, then combines large language models with Inbenta's proprietary NLP, customer-specific lexicons and real-time intent generation to deliver near-zero hallucination rates. The hallucination numbers from raw models remain striking. We discuss a report that shows hallucination rates across different LLMs, including those from companies such as OpenAI, Google, and xAI. Inbenta's approach constrains the model to verified content and context, then layers intent-based NLP on top. In one customer test of 200 questions, they found a single hallucination, caused by a query that fell outside the content boundaries. We also discuss the broader trajectory of enterprise AI. Most deployments today focus on replacing existing activities or functions. The bigger opportunity lies in using AI to imagine entirely new kinds of value. Merlin predicts that hyper-automation, which combines agentic intelligence with robotic process automation, will be the defining shift in the next 12 months. Show notes Find out more about The European Chatbot & Conversational AI Summit: https://europe.customercontactweekdigital.com/events-ccw-uk/agenda-mc/?utm_source=VUX%20World&utm_medium=Media%20Partner&utm_campaign=47758.003_VUX_Social_Post_Agenda&utm_term=&utm_content=&disc=&extTreatId=7634824 Find out more about Inbenta: https://www.inbenta.com Follow Merlin on LinkedIn: https://www.linkedin.com/in/merlin-bise-8277696b Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms Take our updated AI Maturity Assessment: https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript

Where generative AI earns its place in fintech. Lessons from Jaja Finance with Adrian Matei

Most teams are still asking where to use AI. The better question is where not to. In this episode, Adrian Matei, Product Manager from Jaja Finance, breaks down what actually happens when generative AI meets real users and real constraints. The result is far less hype, far more discipline, and a clear view of what it takes to ship AI that works. We talk about Jaja’s chatbot, Airi, which is a RAG-based agentic system with six agents in production, serving customers who often face sensitive financial situations, such as debt. Adrian explains the design philosophy behind Airi’s tone, which was developed in collaboration with Jaja's commercial and communications teams to ensure it matches the brand across every customer touchpoint. A significant part of the conversation focuses on the practical realities of building production-grade generative AI systems. Adrian shares his thinking on when to use generative AI and when deterministic, rule-based approaches are simply the better choice. We also touch on broader themes, including the future of AI at the edge and the emerging role of personal AI assistants. Show notes Discover more about the CCW UK Summit: https://europe.customercontactweekdigital.com/events-ccw-uk/agenda-mc/?utm_source=VUX%20World&utm_medium=Media%20Partner&utm_campaign=47758.003_VUX_Social_Post_Agenda&utm_term=&utm_content=&disc=&extTreatId=7634824 Follow Adrian on LinkedIn: https://www.linkedin.com/in/adriangmatei Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms Take our updated AI Maturity Assessment: https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript

A pulse check on conversational AI in 2026

We hit the floor at the European Chatbot and Conversational AI Summit in Edinburgh with a mic and a simple rule: answer a question, then leave one for the next person. What unfolds is a chain of questions and answers from practitioners at Lebara, Lloyds Banking Group, Uber, ElevenLabs, Carrefour, Virgin Money, Tesco, Jaja Finance and more. The result is an organic, unscripted snapshot of what the conversational AI community is actually thinking about right now. Topics spiral from enterprise AI adoption and the talent crunch, through agentic banking and agent metrics, to some surprisingly personal territory, including AI for mental health support, cardiac research and simplifying the mundane parts of everyday life. Chapters 00:00 Intro 02:24 Chris Miles, Group Product Lead - Chatbots & AI at Lebara 03:12 Kellin Sjoerds, AI Engineer at Essent & Willeke van de Wetering, AI Engineer at Essent 04:39 Andrew Lavis, Chatbot Analyst at Virgin Money 05:08 Mathias Fanschek, Head Retail Strategy & Digital Transformation at Raiffeisen Bank International AG 06:12 Andrei Spiridon, Head Retail Strategy & Business Transformation at Raiffeisen Conversational AI Lab 08:20 Alan Nichol, CTO at RASA 09:44 Adrian Matei, AI Product Manager at Jaja Finance 11:00 Nikoletta Ventoura, Senior AI Conversation Designer at Tesco 11:37 Maria Guermonprez, CX and Product Manager at Spix Industry 12:19 Damien Bird, Cloud Solution Architect at Microsoft 13:12 Gabriele Iuculano, Senior Test Platform Engineer, Schindler Group & Salvatore Raieli, Senior Data Scientist at Oncodesign 14:53 Jared Browne, Group Head of AI Governance & Privacy at Fexco 15:32 Laura Brady, GTM at ElevenLabs 16:20 Laura Ball, Global AI CX GTM and Sales Lead at Zoom 16:50 Sabrina Brunner, Technical Lead at Allianz Direct 18:10 Lorraine Burrell, Conversation Design Lead at Lloyds Banking Group 19:00 Jana Richter, Executive VP AI and Innovation at NFON AG 19:44 Daniel Orenes Ferrandez, Senior Manager - Customer Experience at Uber 21:06 Guillaume Blaquiere, Group Data Architect at Carrefour 21:42 Laura Macleod, Business Applications, Centre of Excellence Lead at Virgin Money 21:57 Kane’s closing thoughts Show notes Find out more about The European Chatbot & Conversational AI Summit: https://theeuropeanchatbot.com Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms Take our updated AI Maturity Assessment: https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript

Your contact centre is sitting on a goldmine of customer insight with Shane Lynn, EdgeTier

Most companies sit on a goldmine of customer insight and do almost nothing with it. Every conversation holds signals about frustration, churn risk, and broken experiences, yet extracting that value can be a slow and manual effort. In this episode, Shane Lynn, CEO and Co-Founder of EdgeTier, unpacks how conversational intelligence is evolving in the age of large language models and what it means for contact centres. Shane explains how EdgeTier can now model metrics such as resolution rate and customer satisfaction across every interaction, rather than relying on the 10–15% survey response rate most contact centres achieve. We also get into the organisational challenge that underpins all of this: contact centres often hold the richest customer insights in the business, but lack the tools or organisational standing to act on them. Shane argues that data-driven contact centres can shift from being reactive cost centres to genuine strategic assets, influencing product, policy and process decisions across the wider business. Finally, we touch on the growing challenge of monitoring AI agent conversations and why automated analysis of those transcripts may actually matter more than it does for human-handled interactions. Show notes Discover more about EdgeTier: http://edgetier.com/ Follow Shane on LinkedIn: https://www.linkedin.com/in/shanealynn Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms Take our updated AI Maturity Assessment: https://vuxworld.typeform.com/to/a26bf9Rr?utm_source=podcast&utm_medium=audio Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript

Testing for trust in the age of AI agents with Amitha Pulijala and Christoph Börner, Cyara

How do you test something that can behave differently every single time? In this episode, we speak with Amitha Pulijala, Chief Product Officer, and Christoph Börner, VP of Engineering, both at Cyara, about one of the most pressing challenges in conversational AI today. The shift from deterministic, scripted IVRs to generative AI agents is changing how conversational systems are built and tested. The episode covers how Cyara approaches testing in a non-deterministic world, including using AI agents to test AI agents, and why traditional ideas like test coverage no longer apply in the same way. Christoph and Amitha share what they are seeing across Cyara’s clients and the wider industry. The discussion unpacks what AI trust means in practice. We talk about hallucinations, model drift, factual accuracy, bias and regulatory compliance. We explore the economics of generative AI. The cost of running it at scale in production and in testing, and how organisations are rethinking where large language models are necessary versus where smaller, domain-specific models are more effective. The discussion also highlights how failure can scale in AI systems, where a single bad response can affect thousands of customers. Show notes Get a free pass to Cyara’s Xchange 2026 in Dallas, TX, using code VUX: https://cyara.com/event/xchange-2026 Find out more about Cyara: https://cyara.com Follow Amitha on LinkedIn: https://www.linkedin.com/in/amitha Follow Christoph on LinkedIn: https://www.linkedin.com/in/christoph-b%C3%B6rner-7721ab147 Follow Kane on LinkedIn: https://www.linkedin.com/in/kanesimms Subscribe to VUX World: https://vuxworld.typeform.com/to/Qlo5aaeW Subscribe to The AI Ultimatum Substack: https://open.substack.com/pub/kanesimms Hosted on Acast. See acast.com/privacy for more information.

• Transcript