June 30th, 2026 | 12 min read
🎧 Listen to the full episode: Spotify | Apple Podcasts | YouTube
Direct answer: Most dental groups treat artificial intelligence as a procurement decision, asking which tool, which vendor, and which price point. Dr Sonia Szamocki, founder and CEO of 01Health and a former A&E doctor, Oxford-trained clinician and BCG strategist, argues that this is the wrong variable entirely. The decisive question is not which AI you buy. It is whether your organisation is designed to absorb what AI delivers. AI readiness is an organisational design problem, not a technology one. The groups that win the next decade will be the ones that start with the problem, build evaluation in from the start, and treat adoption as a partnership rather than a purchase.
The Wrong Variable
There is a question almost every dental group is asking about AI right now, and it is the wrong one.
Which platform. Which vendor. Which price point. The entire conversation has organised itself around procurement, as though the decisive act is the purchase and everything after it is implementation detail.
Dr Sonia Szamocki has spent her career on the other side of that assumption. She began as an A&E doctor, trained in medicine at Oxford, moved through BCG and BCG Digital Ventures, and then founded and now leads 01Health, building clinical AI infrastructure from scratch. She has seen this problem from inside elite strategy and inside the messy reality of shipping a clinical product. And her view is direct: the question that matters is not which AI you buy. It is whether your organisation can absorb what AI delivers at all.
That is not a technology question. It is a question of organisational design. It echoes an argument we have made before, that AI is not simply a better mousetrap, but a technology wave that demands a different kind of business. And it is the question most operators skip straight past on their way to the demo.
"We shouldn't be focusing on the tools that look cool and seem like they probably will have an impact. We need to be thinking about what the problems are first." Dr Sonia Szamocki
The discipline she is describing is unglamorous and almost aggressively basic. Start with the problem. Do you have an efficiency problem, a specific bottleneck, a revenue problem? Only once that is named clearly should the question of AI even enter, and then only if AI is genuinely the right tool for the job. Some of the answers will be AI. Some will be process. Some will be people. The error is starting from the technology and working backwards to a justification.
Why Probabilistic Systems Break Your Governance
The deepest point in the conversation is also the most technical, and Sonia makes it land without jargon.
Traditional software is deterministic. If this, then that. The same input produces the same output, today and a year from today. Governance frameworks across dentistry, and across most regulated industries, are built on that assumption. You validate a system once, you trust it thereafter.
AI breaks that assumption at the root.
"If you ask it to do something ten times, you'll get ten different answers. That creates a real problem when you're building software that has to be reliable and predictable, especially in a healthcare setting." Dr Sonia Szamocki
A system whose output shifts every time, on probability, cannot be governed the way deterministic software is governed. This is exactly why a board-level AI strategy without a governance framework underneath it tends not to work. The implication is one most dental groups have not internalised. The hard work in AI is not the building. The building has become almost trivially easy. Anyone can now spin up a capable tool in an afternoon. The hard work has moved downstream, to evaluation: knowing, on an ongoing basis, whether the tool is still doing the thing you asked it to do.
Sonia puts a number on it. For any build, internal or external, she estimates that between 50 and 90% of the effort sits in evaluation, not construction. That single reframe should change how every operator thinks about both building and buying. A polished prototype proves almost nothing. The question is whether the thing can be monitored, evaluated, and trusted to perform consistently in a clinical setting over time. If a vendor cannot evidence how that happens, the tool is not ready to adopt.
Change Fatigue and the Resistance That Kills Adoption
Most dental teams have been through it already. Digital transformation. Practice management rollouts. Post-pandemic restructuring. By the time a new AI platform arrives, the team is not neutral. It is tired, and it is wary.
Sonia's framing of why people resist is the most useful part of this section, because she does not treat resistance as irrationality to be overcome. She treats it as information.
Resistance usually comes from two places. The first is fear of the unknown, the learning curve nobody has acknowledged out loud. The second is incentives. When staff hear the word efficiency, they hear cost-cutting, and they quietly wonder whether the tool is designed to make them better or to make them redundant. Leadership that fails to address that question directly should not be surprised when adoption stalls.
The fix is to start with the problem the team actually has, and bring them into articulating it.
"Get a blank piece of paper and say, what would you like to happen? What would a magical world look like where this is all fixed for you? Then work backwards." Dr Sonia Szamocki
Do that, and you solve two problems at once. You create the incentive, because the team is solving its own pain, and you create the buy-in, because they are co-authors of the solution rather than its recipients.
The Vendor Carries the Heaviest Load
When the decision is to buy rather than build, Sonia makes an argument that too few vendors will say out loud: the heaviest lifting in AI adoption belongs to the vendor, not the customer. A vendor who deploys a tool and leaves the customer to work out adoption has not sold a solution. They have sold a problem transfer. It is a close cousin of the pattern we examined in why most dental tech products fail, and what actually wins instead.
"If you are hearing that as soon as the contract is signed, the next time you'll see them is in a year at renewal, that's not a good signal." Dr Sonia Szamocki
Responsible vendor behaviour is a partnership that continues well past signature, with genuine skills transfer so the customer's own team can operate and adjust the tool, and a backend that is surfaced rather than buried under the hood. The red flags are the inverse. Do not worry about how it works, just call us if there is a bug. See you at renewal.
From this she draws four questions every dental group should ask a vendor before signing. How will we, and you, check the AI is actually working on an ongoing basis? How will the tool be implemented, and how do we make sure our people can use and modify it? Who actually needs to be involved, which is almost always more people than you think? And the pre-mortem: working backwards from success, what would cause this to fail? A vendor who cannot answer that last one honestly is a vendor to be wary of.
Build on Imperfect Foundations, and Learn From the US
Two closing points sharpen the strategic picture.
First, on data. The common objection is that proprietary clinical data only becomes an advantage if it is clean, structured and consistently governed, which in most multi-site groups it is not. This is the same data layer that private equity is actually pricing when it looks at your practice. Sonia's answer is an 80/20 one. If everyone waited for immaculate data, nothing would ever get built. You can do a great deal with data that is merely in good shape, and the act of trying to build on it is precisely what forces you to clean it up. Do not let the pursuit of perfect foundations become the reason you never start.
Second, on what the UK can learn from a more mature market. The US dental market is two to three years ahead on consolidation and vendor density, and many US groups have been burned by exactly the point-solution sprawl the UK is now tempted by. The lesson is not which tools to buy. It is that the successful mature groups have stopped buying more point solutions and moved toward end-to-end systems that reduce governance overhead rather than multiplying it. The specific decision a UK operator should make now is to resist buying something because it looks impressive and everyone else seems to have it, and instead to select vendors against a clearly named problem.
What This Means For Operators, Founders and Investors
For operators, the implication is that AI strategy begins before any purchase, with the honest articulation of the problem and an honest assessment of whether the organisation can absorb the change. Skip that, and the most sophisticated platform in the market will sit unused. It starts from treating your software as the brain of the business, a tool you own rather than a cost you manage.
For founders and vendors, the bar is higher than a good demo. Evaluation infrastructure, genuine implementation partnership, and skills transfer are the product, as much as the model is. The vendors who treat adoption as the customer's problem will lose to the ones who own it.
For investors, the diligence question shifts. Weight a portfolio company's evaluation discipline and its post-signature delivery behaviour at least as heavily as its model. A tool that cannot be monitored and trusted over time is a liability dressed as an asset.
The organisations that lead the next decade of dentistry will not be the ones with the biggest budgets or the most software. They will be the ones asking better questions about what they are actually building for.
Key Takeaways
AI readiness is an organisational design question, not a technology one. The decisive variable is whether your organisation can absorb what AI delivers, not which tool you buy.
Probabilistic systems break deterministic governance. Because AI output shifts every time, the hard work moves from building to evaluation, an estimated 50 to 90% of the total effort.
Resistance is information, not irrationality. Address fear of the unknown and the cost-cutting fear directly, and bring teams into articulating the problem so they co-author the solution.
The vendor carries the heaviest load. A vendor who disappears until renewal has sold a problem transfer, not a solution. Use the four questions, and the pre-mortem above all, before signing.
Build on imperfect data and learn from the US. Do not wait for perfect foundations, and do not repeat the point-solution sprawl that has already burned more mature markets.
This article draws on the TechDental conversation with Dr Sonia Szamocki, founder and CEO of 01Health. Full episode on Apple Podcasts, Spotify, and YouTube.
About the Guest
Dr Sonia Szamocki is the founder and CEO of 01Health, where she is building clinical AI infrastructure for dental groups and DSOs. Before founding the company, she trained in medicine at Oxford and worked as an A&E doctor, then moved into strategy at BCG and BCG Digital Ventures. That path, from frontline clinical care through elite strategy consulting to building a clinical AI platform from scratch, gives her a rare vantage point on what AI adoption actually requires of a healthcare organisation.
Connect with Sonia and 01Health:
Website: 01health.ai/uk
Sonia on LinkedIn: linkedin.com/in/dr-sonia-szamocki
01Health on LinkedIn: linkedin.com/company/01health
Instagram: @01.health and @aerox.health
About TechDental
TechDental is a strategic intelligence platform for founders, executives, operators and investors shaping the future of dentistry. Through high-level analysis and systems-focused conversations, we explore how AI, governance frameworks and operating model design influence performance, scalability and enterprise value in dental organisations.
www.techdental.com | info@techdental.com
The organisations that lead will be the ones asking better questions.
Copyright Dr Randeep Singh Gill and RIG Enterprises Limited (Company No. 11223423) 2026. All Rights Reserved. TechDental is a trading name of RIG Enterprises Limited, incorporated in England and Wales on 23 February 2018, registered at 1a City Gate, 185 Dyke Road, Hove, England, BN3 1TL. Opinions attributed to named guests reflect the views of those individuals as expressed during recorded interviews. No part of this article may be reproduced without the prior written consent of RIG Enterprises Limited.