Sustainable Technology in Conversation with Mario Michaels

Originally published on LinkedIn.

Oliver had the pleasure of joining Mario Michaels on his Transformational Blueprints Podcast to dive deep into sustainable technology, responsible AI development, and what makes a great architect in today's rapidly evolving landscape. Mario hosts thoughtful conversations with architecture and transformation leaders, focusing on why so many transformation programmes fall short and what we can do about it.

Here are some of the key points from the conversation with Mario, covering everything from the practical realities of cloud optimisation to the critical need for AI governance frameworks—and why human relationships remain more important than technical prowess in architecture roles.

📺 WATCH THE FULL CONVERSATION:

What Makes a Great Architect in Today's World?

"I think in today's world it's got to be critical thinking skills, right? The ability to look through what the vendor or maybe general narrative is and do the classical 'it depends' thing as an architect, right? And go, 'Does this actually work for my industry, in my organisation, or my problem? Or actually, am I being blinded by marketing?'"

The modern architect's most valuable skill isn't technical knowledge—it's the ability to cut through vendor hype and apply contextual thinking to real business problems.

Redefining Sustainable Technology

"Sustainable technology is a little bit overloaded because I think it can mean three things depending on the context... The first is renewable energy and similar technologies, second is what McKinsey describes as 'defensive play' which is: how do I reduce the impact of my technology? The other side is the more "offensive" or proactive play around using technology to solve sustainability challenges."

Oliver breaks down sustainability (partly as per this McKinsey piece) into three critical areas:

  • Renewable energy systems (solar, wind, etc.)

  • Defensive approach to digital (reducing CO2, water consumption, e-waste from tech use) - as per the Tech Carbon Standard

  • Proactive technology (solving sustainability challenges through tech)

The Win-Win-Win Approach

"If you get this right, you're doing win-win-win, right? You're doing things that reduce costs, that reduce risks, and reduce impacts to people and planet... Unfortunately it's been used, I think unfairly, as a divisive sort of tool that goes: there's a camp of people over there and they're activists and hippies, and there's other people over there that are the ones that make money. And that's just too polarising a view that I'd love to challenge."

The most effective sustainability initiatives align environmental benefits with cost reduction and risk mitigation—moving beyond false dichotomies.

Cloud Optimisation: The Low-Hanging Fruit

"Most organisations, if they seriously audited and assessed their cloud spend, they could reduce that significantly. Most stuff's underutilised... If you ask people to go and look at environmental sustainability and cost, people care about that a lot more than just saving money, quite frankly."

Key insight: Adding environmental impact as a driver alongside cost optimisation actually increases engagement and results compared to pure FinOps approaches. Credit to Mark Butcher who regularly highlights this.

The Tech Carbon Standard Recognition

"The UK government have listed it as a useful resource when you're designing sustainable services... Many organisations are surprised when we do this assessment and go, 'You've got 45% of your cost and carbon, quite frankly, in end-user hardware that you're refreshing every two to three years.'"

The open-source Tech Carbon Standard (techcarbonstandard.org) has gained official recognition and reveals surprising hotspots in organisational carbon footprints.

AI: The Critical Need for Architectural Decoupling

"Organizations that get too tightly coupled in their architecture to AI models; that's where, for me, is something I would love architects to start going, 'Do you know what, is this an area of my business that I cannot absolutely be hardcoded into a particular AI model? I need some more decoupling.'"

The geopolitical reality: Changes to OpenAI's filtering mechanisms demonstrate why architectural flexibility around AI models is essential for brand protection and operational resilience. See also some of Oliver's GenAI reference architecture work.

Responsible AI Development

"The trouble with this technology is this phrase that gets thrown around a lot - 'non-deterministic properties of the technology'. Means that it's more human-like... You probably need something similar for AI models, as stupid as that sounds. As in, if they start making mistakes, okay, we need to start monitoring that."

Oliver advocates for treating AI models with governance frameworks similar to employee policies—including progressive discipline and performance monitoring.

The Augmentation vs Automation Debate

"When you use these things for what they're good at, which is processing language, then they blow it out of the park. I think when you try and stretch them too far beyond that human mimicking of language and reasoning, that's when I think people start to get in hot water... It's augment rather than fully automate, I think, is where we are right now."

Practical examples:

The Future of Architecture Talent

"I think almost every person in technology is going to become more of an analyst because they're going to need to understand: where do I do the thing? Where do I get a tool to help me do a thing? Where can I almost fully automate a thing?... I think the architect role will perhaps just onboard more people from delivery management, more from business analysis, than perhaps it does currently from software engineering?"

The architect's role is evolving from technical depth to analytical breadth—becoming the critical decision-maker / evaluator for technology trade-offs.

Architecture Review Boards: Still Relevant for AI

"I would hope that in most organisations now, around AI, the architecture review board is where some of these really quite punchy conversations are happening about AI deployment... There should be conversation happening between the architect(s) and the senior responsible owner for the technology platforms... where is the line on this technology that the organisation is comfortable taking in terms of its brand reputation, regulatory compliance stance, and environmental footprint?"

Despite their reputation, architecture review boards remain the essential forum for AI governance discussions that balance technical capabilities with business risk appetite. Particularly in highly regulated firms.

Quick-Fire Wisdom

Best advice for a newly appointed chief architect:

"Hold your nerve... Don't get too swayed by technology vendors trying to sell you silver bullets."

Career lesson learned too late:

"Human relationships are actually more important than the tech."

Question for fellow architects:

"What are you doing to help your organisation realise and mitigate its environmental impact of tech?"

Massive thanks to Mario for having me on the podcast.

Originally published on LinkedIn.