S1 E20: Designing Systems People Will Actually Trust with Richard Savoie
The Future of Trust with Mat YargerShow Notes
Trust rarely breaks loudly. Most of the time, it erodes quietly, one small doubt at a time. A number that feels off. A recommendation that does not quite match reality. A system that is technically correct, but still ignored. Richard Savoie has spent his career sitting right inside that tension. He started in regulated medical devices, where mistakes can have dire consequences and trust is not optional. In that world, data has lineage, accountability is clear, and failure forces learning. Later, he moved into logistics and transportation, a very different environment, but one facing its own quiet crisis. Labor shortages. Rising complexity. Pressure to automate faster than people are ready to trust the systems doing the work. In this first conversation, Richard shares that his focus on trust does not come only from engineering. It comes from experience. A long professional relationship ended when trust was broken in a way that forced him to rethink how people protect themselves, how they recover, and how they decide who or what deserves their confidence going forward. Today, Richard builds digital twins of logistics networks that help fleets decide how to move, when to electrify, and where automation fits. These systems recommend actions to streamline our orders with UPS, FedEx, Amazon, and more. That is where trust becomes real: if the operators on the ground do not believe the data or the actions they're recommended, they stop listening. If they cannot explain a decision, they override it. So if accountability is unclear, adoption stalls. This episode is about the human threshold where trust either forms or fails. We talk about early environments and how they shape responsibility. We talk about why people trust Google Maps without thinking, and why enterprise systems still struggle to earn that same confidence. And we explore what happens next as AI, EVs, and autonomous systems move from experiments into everyday infrastructure. Chapters: 00:00 – Introduction: Trust Breaks Quietly 03:30 – Childhood, Broken Homes, and Early Trust 10:45 – Moving to Australia and Building a Life 16:00 – When Trust Is Actively Broken 22:30 – Reliability, Systems, and the Light Bulb Cartel 26:30 – Medical Devices and Designing for Certainty 32:30 – Leaving Healthcare and Starting Adiona 37:00 – COVID and the Scaling Moment 43:30 – Why Operators Ignore “Correct” Systems 51:00 – AI, Probabilistic Systems, and the Limits of Certainty 57:30 – Energy Grids, Electrification, and Profit Optimization 1:00:00 – Autonomous Delivery and the Last Meter Problem 1:12:00 – Being Known Well
More from The Future of Trust with Mat Yarger
In this episode, we sit down with Kazuna Yamamoto, a founder, venture builder and activist based in Berlin. Born in Hong Kong, raised in Singapore and Japan, and shaped by years in Chile, Kazuna learned early that trust is something you build from scratch when you arrive somewhere with no safety net. The conversation moves from a viral petition she launched at 21 against a Japanese tabloid, which made her a national target, to her one-way ticket to Santiago and a fintech startup frozen by COVID. Kazuna and Mat dig into how growing up in multicultural environments changes who you trust by default, why dance taught her that vulnerability is a form of trust, and the moment a friend’s actions cost her credibility with people who had vouched for her. As Mat puts it, trust is a one-to-many relationship. Kazuna also walks through Voice Up Japan, the organization she runs, and its Femicide Project, including an AI-built femicide tracker and an agent that matches domestic violence victims with pro bono lawyers. She closes by naming the human skill she thinks matters most as AI takes on more: empathy. In her words, “I just hate the world a lot, but I love the world a lot.” If you enjoy the episode, don’t forget to subscribe and leave a review!
In this episode, we sit down with Peter Keutgens, an actuary and data scientist who spent more than two decades at WTW in the UK, where he went from quant researcher to leading the firm's quantitative team. Peter grew up in Turnhout, a provincial Belgian town close to the Dutch border. He was the quiet kid in the corner building with Lego. A maths teacher who doubted him early on taught him something he still uses: a title gives someone weight to voice an opinion, but it is still only an opinion. The conversation starts with the manager who pulled Peter out of "The Quant Corner" and taught him to communicate. It moves on to the second career he built after redundancy, which combines actuarial data science, blockchain and his love of flying gyrocopters. Peter explains why what really makes an aircraft fly is "money and paper," and how handwritten maintenance records erode trust when an aircraft changes hands. He also explains why his Aviation Trust Network should be a nonprofit standard. Mat and Peter dig into digital twins and data ownership, including how to keep a pilot's personal flight data apart from an aircraft's permanent record. They then unpack Peter's modelling of costs under the EU Emissions Trading System (ETS) and of the supply of sustainable aviation fuel (SAF). One finding stands out: if carbon captured from the air earns full ETS credit, that CO2 will sell for at least the price of an allowance. That makes synthetic jet fuel more expensive than many people expect. If you enjoy the episode, don't forget to subscribe and leave a review.
In this episode, we sit down with Oneza Zaim, a carbon markets analyst at Planet2050, a climate fintech company financing high-integrity CO2 removal projects and opening carbon assets up to retail investors for the first time. Born in Abu Dhabi and raised across three very different environments, Abu Dhabi, Pakistan, and now Berlin, Oneza traces her understanding of trust back to a childhood spent learning other people's languages just to make herself understood. The conversation moves from her training as an industrial auditor, through a site visit where she walked into an unmarked chemical vapor leak and had to call a halt on the spot, to her current work vetting carbon dioxide removal projects at Planet2050, where her team has screened more than 250 projects across different CDR technologies. Oneza and Mat dig into what actually makes a climate project bankable: not a checklist, but a matrix of who is carrying which risk, how dependent a project is on subsidies that could disappear, and whether the story on paper matches what you find on the ground. As Oneza puts it, "integrity does not mean that you do not have weaknesses, it means you know where they are and what you're doing about them." They also unpack why projects tied to an existing supply chain, like coffee waste turned into biochar, tend to earn investor trust faster than nature-based projects like mangroves, even when the latter matter just as much. If you enjoy the episode, don't forget to subscribe and leave a review.
In this episode, we sit down with Andrew Forson, President of DeFi Technologies and a serial entrepreneur working across asset management, cryptocurrency, and blockchain infrastructure. Raised in a small blue-collar town outside Montreal by two Ghanaian immigrant physicians, Andrew traces his approach to trust in institutions, markets, and government back to a childhood built on self-reliance and hard work. The conversation moves from wrapping digital assets into regulated investment products, to QuickPick's blockchain-powered remittance platform for underbanked users, to the moment a trusted co-founder tried to sell stolen IP behind his back. Andrew and Mat dig into what he calls the real break in global trust: sovereign debt. With nearly all of the world's 195 countries carrying unsustainable debt loads and politicians facing zero accountability for it, Andrew argues "you are certainly a shareholder in the country's outstanding debt" — whether you realize it or not. They also unpack why small, disciplined economies like Switzerland, Singapore, and Rwanda earn outsized trust, and what it would take to give the world's poorest access to fair, inflation-proof investment. If you enjoy the episode, don't forget to subscribe and leave a review.
From investigating forged documents on Estonia's post-Soviet border to founding ComplyOnce, Raido Saar has spent his career chasing the same question: how do you know who's really on the other side? Along the way, he helped shape Estonia's X-Road digital government infrastructure and built one of the first biometric passport identity solutions in Europe. In this episode, Raido sits down with Mat Yarger to unpack why the world spends $500 billion a year on anti-money-laundering efforts yet recovers just 0.02% of illicit funds, how outdated identity checks let one scammer open five fraudulent accounts using the same stolen face, and how Raido himself nearly got scammed on a fake podcast call. He also shares how LinkedIn and its verification vendor lost his identity data twice, and makes the case for a single encrypted identity key that could replace the passwords, selfies, and stolen accounts fueling today's fraud epidemic. It's a candid, wide-ranging conversation about why our systems for verifying trust are still built on 1990s architecture, and what it will actually take to fix them. If you enjoy the episode, don't forget to subscribe and leave a review.