John McMullan says he has worked in conversational AI for more than fifteen years.
“I've been in conversational A I somehow, some way, fifteen plus years.”
Person
John McMullan
1 episodes · 19 quoted claims
John McMullan argues that AI, unlike traditional software, is a continuous investment that needs ongoing tuning rather than a one-time installation.
“A I is the continuous learning cycle, a continuous investment where the system is going to uncover new edge opportunities, need constant tuning and tweaking.”
John McMullan says he meets with dozens of customers every week, which informs his view of common AI adoption problems.
“with the privilege I have of sitting down with, you know, dozens of customers every week and understanding what they're working on”
John McMullan says AI steering committees consider ethical use cases but are not the implementers who build the systems.
“Those are the people that step back and say, what are the ethical use cases?”
John McMullan says an after-call wrap-up AI feature fills out the call disposition when a customer call ends.
“So when a call concludes between a customer and a frontline rep, it fills out the disposition.”
John McMullan says that at one site he visited, agents pasted call transcripts into a free ChatGPT instance to check the AI-generated notes against their own.
“those agents were actually taking the call transcript, copy and pasting it into a free instance of chat G P T and then measuring those notes back against their notes.”
John McMullan says the agents' distrust stemmed from a non-mandatory training that did not explain why the new tool appeared on their desktops.
“it was a non mandatory training. They didn't understand why this new widget showed up on their desktop.”
John McMullan cites coaching reports delivered only every two to three weeks as an example of a customer-experience friction point AI could address.
“Why is a frontline, you know, agent getting a coaching report every two to three weeks that's cobbled together?”
John McMullan lists long hold times and phone menus as customer friction points where AI projects should start.
“wait a minute, why am I waiting on hold for 18 minutes? Or why am I pressing three for billing”
John McMullan says one of the hottest companies coming out of Y Combinator works on AI agent evaluation.
“I think one of the the hottest companies coming out of Y Combinator”
John McMullan says that asking the same question to a large language model can produce different responses each time, which is a risk for brands.
“if everybody listening went to your favorite L L M and we all asked it the same question, we may all get different responses.”
John McMullan says simulating conversations and edge cases lets teams test that an AI follows policy consistently before release.
“make sure you can simulate all of those conversations, those edge cases, put them through and ensure that you're having consistency on every interaction”
John McMullan says an AI agent's data readiness includes wiring in back-end systems such as CRM, ERP, billing and scheduling.
“Making sure that your back end systems, your C R M, your E R P, your billing systems, your scheduling systems, all are wired in.”
John McMullan recommends that experienced frontline reps, who handle hundreds of conversations weekly, test AI agents because they know the edge cases.
“making sure your frontline reps who already handle hundreds of these conversations every week go and actually test it.”
John McMullan says frontier models from companies like OpenAI and Anthropic are moving toward long-term planning of multi-step processes such as loan applications and insurance claims.
“it is the long term thinking and planning for multi-step procedural operations. Think of like loan applications, right?”
John McMullan defines the interaction intelligence layer as a layer that takes in every customer touchpoint and sits between core business systems and AI agents.
“sits in between your core systems and your A I agents. That is the interaction intelligence layer.”
John McMullan cites a saying, which he attributes to another author, that IT departments will become the HR of AI by managing many AI systems.
“I T is going to become the H R of A I.”
John McMullan says companies likely already hold tens of thousands of successful customer calls that can show where and how to build AI.
“you are probably already have tens of thousands of great customer calls where you know it went right”
John McMullan gives a hypothetical in which ten people are moved off an automated process but twelve end up maintaining the AI, to illustrate ongoing maintenance costs.
“I moved ten people from this process that we've automated, but now I got twelve people working on maintaining the A I.”