Ross Barnes was a global CTO at one of the largest agencies in the world before he understood why work felt the way it did. Years into a career most people would call successful, a deep dissatisfaction kept surfacing. Emotional overload would hit at least a couple of times a year, and it hurt every time. He had been to therapy before. None of it had named the thing he was carrying.

A psychotherapist finally suggested he get evaluated for ADHD. He did. The diagnosis came back, and so did one for autism. The world he had been navigating was not built for the way his brain worked, and once he could see that clearly, he could no longer unsee it.

Ross is the Founder of The Galahad Group, an AI platform that builds what he calls cognitive scaffolding for individuals and organizations. The company was not designed in a strategy session. It was designed around one person's brain, then opened up to everyone else.

In this episode of Lead with AI, Dr. Tamara Nall speaks with Ross about how a late-in-life diagnosis became the architecture of a company, why his platform refuses to automate work that belongs to humans, and what happens when AI is built to make people more human instead of less.

 The First Customer Was the Founder   

Galahad's operating system was built around how Ross actually thinks, not how he was supposed to think. His personal version of the platform holds him accountable for the things his brain misses. It nudges. It prompts. It checks his headspace before pushing him into something demanding. Ross says the work he does on himself, what he calls the gym of the mind, matters as much as the work he does on the company.

Once the system worked for him, the question became obvious. If the platform could make one neurodivergent brain happier, healthier, and more functional inside corporate work, what could it do for everyone else? The answer became Galahad.

 Four H's That Decide What Stays Human   

The framework at the center of the platform is called ikgiAI, named after the Japanese concept of finding a reason for being. Ross adapted it into a diagnostic that walks through every workflow inside a company and sorts each task. Some belong to people. Some belong to agents. The decision runs through what Ross calls the four H's of humanity.

Does the task need someone to be here, present in the moment? Does it need hunger, the resilience to push through something difficult? Does it need heart, the empathy that builds trust? Does it need a hunch, the instinct that catches what data cannot? If any of those four are part of the work, Galahad will not touch it.

The line is non-negotiable. When a paying client asks the company to build something the framework flags as inherently human, the company refuses. Ross's most-used phrase with his team is that just because we could does not mean we should.

 A Live Keynote and a Working Agent in Real Time   

Ross was on stage at a large European publisher delivering a keynote when he ran a competition with the audience. Each attendee submitted a use case for how agentic AI could solve a real problem in their work. Ross judged the entries live, picked a marketing challenge focused on delivering the right content to the right consumer in the right voice, and built the agent inside his platform while the audience watched. By the end of the session, the agent worked. The phrase "holy smokes" got said more than once.

That moment captures the bet behind Galahad. The platform turns a real workflow problem into a deployable agent fast enough that the audience watching can see the loop close. Not a slide. Not a roadmap. A working solution.

 Catching the Shadow AI Hiding Inside a Global Brand   

One of Galahad's beta tools tackles a problem most companies do not realize they have. Employees across every function are quietly pasting sensitive material into their own ChatGPT accounts. The data leaves the company. Nobody notices. Galahad's shadow AI detection inserts canaries into company content, identifiers that can be queried inside ChatGPT later. If a canary surfaces, the company knows exactly which document leaked and where it went.

A large fast-moving consumer goods client deployed the system. The result was not a crackdown on AI use. The opposite. Once leadership could see where data was actually going, they were able to give staff better internal tools instead of locking everything down.

The shadow stopped being a problem when the lights came on.

 A Board of Advisors That Actually Disagrees With You   

Ross spent an hour interviewing one CEO and asked for the executive's job description and a week of his calendar. He fed everything into the ikgiAI diagnostic, and the framework returned a complete picture. Meetings the CEO should not have been attending. Meetings that should not have existed at all. Tasks that should have been handled by an agent before they ever reached a human. Galahad then built a suite of agents around the executive, including one designed to disagree.

That last piece matters. Most leaders cannot afford to hire someone whose entire job is to push back. Few people can absorb that kind of friction from a colleague without it eroding the working relationship. An agent built to challenge can hold the line without the social cost. The CEO gets the dissent he needs without the resentment that often comes with it.

 The Better We Know Ourselves, the Better We Know Each Other   

When Dr. Nall asked Ross why he was so committed to building AI that protects the human parts of work; his answer pointed back to the diagnosis. Once he understood how he thought, he could understand how he communicated. Once he could communicate, he could help other people communicate with him. The pattern repeated. The more clearly he could see himself, the more clearly he could see everyone else.

That insight runs through every layer of the platform. Galahad does not aim to make people more efficient. It aims to make people more themselves at work. The agents handle the meetings that should not exist, the tasks that drain energy without producing value, and the cognitive overhead that gets in the way of doing the work that actually matters. What is left over is the human work. The presence. The resilience. The empathy. The instinct.

 Eighteen Months Until an Avatar Is Sitting in Your Meeting   

When Dr. Nall asked Ross for his boldest prediction; he did not soften it. Within 18 months, he expects digital avatars to be attending meetings in our place, acting on our behalf, and reporting back. Not in five years. Not in a decade. Inside two.

It is the kind of forecast that sounds reckless until you walk through what Galahad already does. Agents that absorb meetings the executive should not have been in. Agents that bring opinions and challenge decisions. Agents that monitor data flow, hold people accountable, and remind them they are human. The avatar in the room is the next obvious step. The infrastructure to support it is already being built.

For founders and leaders ready to figure out which parts of their work should stay human and which parts can be supported by agents. Take the ikgiAI diagnostic to see what the platform can do for your role and your team.

For more conversations with the leaders building the future of AI, subscribe to Lead with AI on your favorite podcast platform.

Follow or Subscribe to Lead with AI Podcast on your favorite platforms - Website: LeadwithAIPodcast.com | Apple Podcasts: Lead-with-AI | Spotify: Lead with AI | Podbean: Lead-with-AI-Podcast | YouTube: @LeadwithAIPodcast | Facebook: Lead with AI | Instagram: @LeadwithAIpodcast | TikTok: @LeadwithAIpodcast | Twitter (X): @LeadwithAI

Follow Dr. Tamara Nall - LinkedIn: @TamaraNall | Website: TamaraNall.com | Email: Tamara@LeadwithAIPodcast.com

Follow Ross Barnes (Founder, The Galahad Group) - LinkedIn: @rossbarnes

Follow The Galahad Group - LinkedIn: @The-Galahad-Group | Email: hello@galahadgroup.co.uk