From Bots to Brains: Navigating the New Frontier of Conversational AI - IntelePeer

Episode 9: From bots to brains: Navigating the new frontier of conversational AI

Paula Rivera: Welcome to the AI Factor, where business meets AI. I’m your host, Paula Rivera, and in today’s CX Explained episode, we are diving into the 2025 Opus research conversational AI Intelliview report with the one and only Derek Top, Principal Analyst and Research Director of Opus Research. We’ll explore how the conversational AI landscape is evolving from flow-based bots to gen AI-driven agents, platform maturity, and what really separates leaders from laggards. If you are evaluating a provider or just trying to keep up, this is the episode for you, Derek, welcome.

Derek Top: Thank you. Good to be here.

Paula Rivera: Great to have you. Before we dive into the heart of your report, I have to ask... yesterday, the White House released its winning the Race America’s AI action plan. According to CNN, the plan has three pillars: Accelerating innovation, building out AI infrastructure in the US, and making America hardware and software the standard platform for AI innovations built around the world. Have you had a chance to review the plan yet and do you have any initial reactions?

Derek Top: Yeah, I’d say I’ve reviewed it. I’ve not read it word for word all the way through. I know it came through yesterday, and yeah, I think it is a very interesting take on where we’re going with AI, both as a country from the US and also from a global perspective. So it aligns with a lot of the industry hopes for accelerating the pace of adoption and using AI as a strategic asset. There is a lot of hope for what’s next in terms of how the US can be a leader in the AI race, if you will.

Paula Rivera: I concur completely, and I was of the same mindset as most of the leaders in the room yesterday thinking this is all very positive and definitely needed.

Derek Top: This plan pushes towards streamlining regulations and building data centers faster for exporting AI software to the world in terms of semiconductor manufacturing. It also addresses potential issues around open-source models and open models.

There’s some language about training models that use, "quote-unquote," around bias, and there is potential for challenges here, but overall, I think the goals around making it easier for tech companies to develop, test, and deploy AI is a good thing.

Paula Rivera: Let’s get into the report, which I found very thorough. One of the big shifts you highlighted is the U.S. moving from flow-based bots to prompt-based gen AI-powered agents. Why is this such a significant turning point?

Derek Top: The introduction of generative AI and large language models has changed the landscape of conversational AI. Traditionally, intelligent assistance and bots were focused on automated self-service to help reduce call volumes and repetitive tasks using deterministic dialogue design. With the role of generative AI, the design becomes more flexible, allowing bots to better understand user input, offering a more human-like interaction.

Paula Rivera: How are enterprises navigating between the predictable and the flexible?

Derek Top: Many organizations have relied on IVRs which have been limited in terms of functionality and business outcomes. Now, generative AI is creating exciting opportunities for businesses to save money and improve customer interactions through more flexible bots. This allows for tasks such as transaction handling or appointment scheduling to be addressed in a more dynamic manner.

Paula Rivera: People still complain about customer service bottlenecks. Is this due to companies being slow to adopt modern technologies?

Derek Top: Yes, there's a limited understanding from the bot perspective. People have been frustrated by older technologies, and as expectations change, organizations are now looking to adopt AI-powered solutions that can better respond to user needs and frustrations.

Paula Rivera: Let’s unpack the distinctions you make between pragmatists and true believers regarding the evolution of gen AI.

Derek Top: The distinction allows us to categorize companies based on their approaches to technology; pragmatists want practical solutions and to leverage past technologies, while true believers are more forward-thinking. We see both as important in the evolution of conversational AI, and there's a push for businesses to transition fully into the AI realm while still leveraging existing technologies.

Paula Rivera: It makes sense that vendors who straddle both strategies are likely to succeed with larger customer bases.

Derek Top: Exactly. As technologies continue to evolve rapidly, organizations should partner with solution providers to understand how to implement AI strategically and effectively.

Paula Rivera: Can you tell us about the generative AI platform maturity framework mentioned in your report?

Derek Top: We identify five key pillars of platform maturity: Agent orchestration, tools for use cases, knowledge management, observability to monitor control, and evaluation and trust concerning security protocols. Each pillar represents critical aspects companies should consider in implementing AI.

Paula Rivera: Trust and security are crucial. How can vendors and organizations assess these areas beyond marketing claims?

Derek Top: Organizations should form task forces to tackle trust and security by bringing in different stakeholders to identify relevant security layers and compliance needs. This proactive approach ensures that AI can be safely and effectively integrated into operations.

Paula Rivera: What red flags should enterprises be aware of when evaluating conversational AI solutions?

Derek Top: The main red flags involve aligning solution capabilities with business goals and being able to keep up with evolving technologies. If an organization is unable to adapt its business processes to leverage new capabilities, it may indicate a serious risk.

Paula Rivera: Let’s talk a bit about pricing models. Do you have advice for matching pricing to business strategy?

Derek Top: Pricing models vary based on needs. Organizations typically use models based on agent counts or consumption-based pricing based on interactions. Leverage solution providers to find a pricing structure that fits your goals and to understand potential spikes in costs during peak times.

Paula Rivera: As we near the end of our segment, I'd like to do a rapid-fire round. What’s one myth about conversational AI that drives you crazy?

Derek Top: The belief that conversational AI is not trustworthy. We're reaching a point where more use makes AI more beneficial, though continuous monitoring is necessary.

Paula Rivera: Favorite use of AI in your daily life?

Derek Top: I use AI to consolidate and summarize information, making my life easier.

Paula Rivera: If you could automate one part of your job, what would it be?

Derek Top: I would automate tasks like scheduling and follow-ups to leverage AI as a powerful assistant.

Paula Rivera: Thank you, Derek! This has been an incredible conversation. Your insights on the 2025 Conversational AI Intelliview report are enlightening. For those exploring platforms, be sure to check out the report for actionable intelligence on navigating the AI-powered customer experience landscape.