“What happens after we’ve taught the robots to talk?”
That’s the question I posed two weeks ago, as conversational AI continues to trend towards: problem solved!
This week, my colleague, friend and zen-master of AI-driven transformation, Dan Aslet, painted a vivid picture: they wander the corridors of your tech stack, bumping into locked doors!
With Gen AI, you can create wonderfully engaging and natural text and voice interactions. But the quality and depth of those conversations now exceeds the capability of your systems and APIs. As Dan put it:
“AI Agents = More with Less. But…that promise collapses unless your agents can reach your systems”
Your next-gen IVR or chatbot can have a great conversation, but it’s limited by the information it can (or can’t) access, and what fixes, changes, or updates it can actually make.
The classic solution is to build more – or better – APIs. But every new API update needs backing into the AI config in every single IVR and chat app you build.
And that’s a problem. Because you’re going to be building a lot more next-gen IVR and chatbots – let’s call them AI Agents, because that’s what they are – on your journey towards an AI first contact center.
And you’re going to be building a lot more APIs to support them.
It’s a classic many-to-many conundrum.
Many more AI Agents.
Many more APIs.
Everything needs to talk to (nearly) everything.
But there’s a solution. Model Context Protocol (MCP). Back to Dan:
“APIs function like a speakeasy—where only the secret knock gets you in… MCP turns enterprise access from a secret knock on a locked door into a bright menu in the window, where agents can browse capabilities before stepping inside.”
Welcome to the age of loosely-couple AI agents and services that can discover each other’s capabilities and leverage them without explicit, fragile and expensive integration. It’s hard to underestimate the breadth of possibilities – and value – that can be created when ‘everything can talk to everything without needing explicit integration’.
MCP is still early, but it’s gaining rapid adoption and it’s pretty clear where the puck is headed. If you haven’t considered – or been asked about – how MCP fits into your AI transformation and customer service strategy, it’s only a matter of time.
Dan’s article is definitely worth grabbing a coffee and a comfy seat for – check it out here: MCP x APIs. His substack is new, but definitely worth a subscribe. There’s gold inside that marvellous brain of his and I’m looking forward to hearing more of it.
There’s an interesting connection between the MCP x APIs post and the need for looser coupling to make complex systems robust:
“Organizations need to build in slack and buffers – what Perrow would call creating “loose coupling” – even at the cost of some efficiency, to prevent cascading failures”
That’s a quote from Michael Fisher’s article: The Possible, the Practical, and the Profitable that I shared last week.
These intersecting perspectives on MCP: from me, Dan, and Fish (yep, that’s what we call him!) is a great example of why we’re on track to expand – and rebrand – this weekly email for the new era of AI.
AI, CX and IT are no longer separate disciplines. Large (and small) Business to Consumer (B2C) companies need leaders – and practitioners – who can connect the dots between these three pillars of modern customer service and business transformation.
What can AI do? How can you harness it? Disrupt with it? And avoid being disrupted by it?
What customer experiences does it make possible? How will it change customer journeys and expectations?
What IT & digital infrastructure do you need to make this happen? How can you get infosec and compliance on board?
These three themes cannot sit in isolation. The companies, teams, and leaders of the future will seamlessly integrate them.
That’s a lot to keep on top of, but we’re here to help.
Kerry
PS: If you want a more regular dose of insights, follow or connect with me on LinkedIn for regular posts on conversational AI, mindset, and egg juggling, among other things!
PPS: You are building with GenAI right now, aren’t you? If not, what’s stopping you? Check out our blog on Gen-AI blockers, or sign up for a complimentary Strategy Workshop to help you get started.
If someone forwarded this to you, please subscribe yourself for weekly insights that’ll make you think differently about your IVR, voice, and chatbots.
Helping you get maximum ROI from conversational AI — whatever the platform.