The Bespoken AI Show

What does it take to build AI agents that are both capable and reliable in the real world?

In Episode 7 of The Bespoken AI Show, John Kelvie sits down with Alan Nichol, CTO and Co-Founder of Rasa, for a wide-ranging conversation about the evolution of AI agents and what their growing capabilities mean for businesses.

They discuss the shift toward speech-to-speech models, the challenges of bringing voice AI into enterprise contact centers, and the balance between agent autonomy and deterministic controls. They also explore enterprise governance, AI risk, and why better evaluation methods are key to building agents organizations can trust.

A conversation about where AI is headed—and what it will take to make it work at scale.

Guest: Alan Nichol, CTO & Co-Founder of Rasa
Host: John Kelvie, CEO & Co-Founder of Bespoken

AI is changing what’s possible in conversational experiences — but building effective AI agents takes more than a better model.

In this episode of The Bespoken AI Show, John Kelvie sits down with Braden Ream, CEO of Voiceflow, to talk about the evolution of conversational AI and what it takes to bring AI agents into real-world enterprise environments.

They discuss the shift from traditional chatbots to agentic systems, the challenges of enterprise adoption, AI hallucinations and governance, the role of deterministic logic, and how AI agents could reshape customer experience beyond traditional contact center interactions.

A conversation about the technology, business decisions, and new ways of thinking shaping the next generation of conversational AI.

In this episode, John sits down with Andrew Freed, Distinguished Engineer and Inventor at IBM, to talk about what it really takes to build AI systems that work in production.

They explore why testing becomes even more important as AI systems grow more autonomous, how teams should define what “good” looks like, and where guardrails, governance, and continuous monitoring fit into the development process.

The conversation also covers the challenges of voice AI, using LLMs as judges, choosing the right evaluation metrics, improving AI systems after launch, and why a smaller set of high-quality test cases can be more valuable than thousands of poorly chosen ones.

A practical discussion on building, evaluating, and improving reliable AI in the agentic era.

Voice AI has been through a lot.

Amy Stapleton, Senior Analyst at Opus Research, joins John Kelvie to look back at the early days of Alexa and conversational AI, what went wrong with the first wave of voice technology, and how LLMs changed the equation.

They talk about where voice AI and AI agents stand today, the changing role of CCaaS providers, trust and testing, and whether most customer service could eventually be automated.

Then they look ahead: Will smart speakers make a comeback? What happens to contact centers when AI agents become capable of handling more complex work? And how close are we to talking naturally with robots?

A conversation about where voice AI came from, where it is today, and where Amy and John think it could go next.

In this episode, John sits down with Emerson Sklar, Chief Evangelist of Amazon Alexa, to talk about Alexa, the evolution of voice AI, and where the technology is heading.

They explore the return of the personal AI assistant, more natural voice interactions, multimodal experiences, and how AI is changing what developers can build.

Emerson also shares personal experiences using Alexa and new AI-powered developer tools.

And if AI can build so much for us, what role do developers still play?

A conversation about Alexa, voice AI, developers, creativity, and what becomes possible when technology stops being the biggest limitation.

00:00 Intro04:05 Why Voice AI is getting better at understanding us08:44 Alexa+ and the vision for a Personal AI Assistant12:16 How Alexa uses your data to become a better assistant17:18 How voice changes the TV experience - FireTV28:57 The Future of the Alexa Developer Community37:22 Do developers still have a role in the Age of AI?

In this episode of The Bespoken AI Show, we break down three big stories shaping the Voice AI landscape:

• OpenAI’s move into dedicated voice hardware
• Google and Five9’s growing partnership, and
• Gartner’s latest Conversational AI Magic Quadrant.

We discuss what OpenAI’s new device says about the future of real-time voice, whether Google could acquire Five9, why Gartner’s rankings raise some big questions, and the importance of having a direct channel to the end customer.Plus, we make a few predictions about where the industry is heading next.

In our very first episode, we dive into some of the biggest stories shaping the Voice AI industry.

We discuss Salesforce's $3.6B acquisition of Fin and what it means for the future of AI-powered customer service, Apple's latest Siri announcements at WWDC, Amazon Connect's continued push toward no-code customer experience design, and ElevenLabs' new Procedures feature for AI agents.

More importantly, we go beyond the headlines to explore what these announcements reveal about where Voice AI is headed, how enterprise adoption is evolving, and which trends we think will have the biggest impact over the next few years.

Join us as we break down the news, share our perspectives, and discuss what it all means for the future of Voice AI.

Topics covered:

• Salesforce acquires Fin for $3.6B

• Siri AI

• Amazon Connect's new UI vision

• ElevenLabs Procedures

• What we learned this week