Talking HealthTech podcast recap — Episode 611: Bold Upgrades, Zero Downtime: How Smart Labs Are Modernising Pathology
Featuring Dan Burke, Managing Director APAC, Magentus, and Ben Richardson, CEO, Labflow
Key takeaways
- Australia processes half a billion pathology tests a year, yet the average pathologist is aged 50 to 55 with a 16-year training pipeline behind each replacement.
- Big-bang LIMS replacements are expensive, risky, and slow to deliver value. Incremental modernisation layers new capability on top of existing systems.
- Labflow’s LabConductor integration layer connects data sources without becoming the master system, enabling labs to modernise without disrupting 24/7 operations.
- AI’s highest-value role in pathology today is workflow and data integrity, not diagnostic decision-making.
- The convergence of workforce pressure, AI maturity, and genomic sequencing costs falling below $100 is creating the foundations for a shift from reactive to preventive healthcare.
Pathology doesn’t get much airtime. It runs in the background and when it works well, nobody thinks about it. But roughly 70% of clinical decisions depend on a pathology result, and the system producing those results is under more pressure than most people realise.
On a recent episode of the Talking HealthTech podcast, Magentus Managing Director Dan Burke and Labflow CEO Ben Richardson sat down with host Peter Birch to talk about what’s actually happening inside pathology laboratories, and what needs to change. The conversation covered workforce pressure, legacy technology, AI, consumer-driven diagnostics, and the genomics cost curve that’s quietly reshaping the future of personalised medicine.
The picture they painted is urgent but optimistic. And their central argument, that labs should modernise incrementally rather than wait for a big-bang replacement that may never come, has implications well beyond pathology.
Why is Australia’s pathology workforce under pressure?
Dan Burke jumped straight into the pointy end of the situation: Australia processes half a billion pathology tests a year, yet the workforce overseeing that volume is aging, with the average pathologist aged between 50 and 55. And it takes roughly 16 years to train a replacement.
“One of the key issues our customers and the pathology industry is dealing with is how to manage that workforce challenge,” Burke said. “The urgency is pretty pressing now in terms of how many people will retire from the pathology workforce over the next two, three, four years.”
Alongside the workforce squeeze, demand for more complex diagnostic testing such as genomics, anatomical pathology and functional panels continues to grow.
How Covid changed consumer expectations in diagnostics
Ben Richardson’s entry into pathology came from an unexpected angle. During the pandemic, large-scale labs had to engage directly with consumers for the first time, including walk-in PCR clinics, QR code check-ins, and results delivered by SMS.
“For the first time ever, at scale, pathology labs had to engage directly with consumers,” Richardson said. “Usually it’s a referral, maybe you see the lab, but it all goes through a doctor. This was a fundamental shift in the customer experience.”
That shift didn’t reverse when the pandemic ended. It revealed a latent consumer appetite for accessible, transparent diagnostics, particularly in out-of-pocket markets across the Asia-Pacific region. Richardson, who has since expanded Labflow into India, Malaysia, and Singapore, sees this trend accelerating.
“Particularly in India, it’s very consumer-driven. The majority of pathology is out of pocket. And that is where we’re seeing the strongest demand from consumers for maximum tool usage at their disposal.”
Patients increasingly expect the same kind of accessibility and immediacy from diagnostics that they get from every other service in their lives, so the industry needs to keep up.
What is the fragmentation problem in pathology?
A theme running throughout the episode is fragmentation. Rather than a shortage of technology, the problem is a lack of connection between the technologies that exist.
Burke was direct about the challenge. “Over 30 years since Magentus has existed, there’s been no shortage of healthcare technology. What’s becoming more of a problem is the fragmentation and the fact that pathologists and other users in labs are having to switch between multiple systems to manage a case from end to end.”
The consequences include slower turnaround, more manual checks, greater risk of error, and delays that ultimately reach the patient. And much of the underlying infrastructure dates back decades.
“Pathology is really interesting,” Burke said, “because when it works really, really well, you don’t see it. The services we provide are at their best when no one notices they’re there because they just work seamlessly in the background. When things do fragment, though, when you’re putting more and more pressure onto systems, some of which have their origins back in the ‘90s, you do start to see things creaking.”
Why big-bang LIMS replacement doesn’t work
The natural instinct might be to rip everything out and start fresh. Burke and Richardson both pushed back hard on that idea.
“It is tempting sometimes to think we’re just going to throw the whole thing out and start again,” Burke acknowledged. “But pathology isn’t a system you can just turn off. It’s a 24/7 system dealing with hundreds of millions of tests. The workflows are very ingrained in how people do their work.”
His conclusion was blunt: “Replacing everything all at once is just too risky. It’s expensive, and it takes a long time to deliver value. Big-bang replacements often fail.”
The alternative, and the strategic rationale behind Magentus bringing Labflow into the family, is incremental modernisation. Rather than replacing legacy LIMS platforms wholesale, the approach is to layer new capability on top of what already works.
Richardson described the philosophy behind Labflow’s product suite: “We created a whole bunch of different products, about a dozen, that you plug on top of your existing ecosystem. That means you can start delivering new value, generating new revenue, reducing human error, and improving patient outcomes very incrementally and in a much more economically appropriate way.”
How does LabConductor solve laboratory interoperability?
Central to Labflow’s model is an integration layer, LabConductor, that connects data sources without trying to become the master system.
“Whatever incremental gains you need, it needs to be very oriented around the principle that we will not be the master holder of that ID,” Richardson explained. “Let the other systems continue to hold that, but make sure everything links up and maps dynamically throughout the experience.”
In practical terms, that means plugging into databases, instruments, or file systems, manipulating data mid-flight, verifying it, and routing it to where it needs to go, all without requiring existing systems to change how they manage patient identifiers, order numbers, or barcodes.
“That is the difference between having to double-barcode things, double-scan things, and significantly increase the labour and cost and opportunity for something going wrong,” Richardson said. “Architecting that way is absolutely crucial for being able to scale with multiple systems already in place in the labs.”
What role does AI play in pathology today?
The conversation about artificial intelligence put the buzz into helpful context. Burke framed AI as a support act in pathology, not necessarily the headline. “It’s important clinicians stay in charge of decisions.”
Richardson agreed, but was specific about where AI is already delivering value: not in replacing diagnostic judgment, but in catching the small, frequent inconsistencies that slow labs down.
“We take a report and look for all the different results, all the different interpretations, and make sure that lines up with all the clinical history across multiple databases, before the report goes out to the patient,” he said. “Equally, we look at request forms and all the data coming into the lab and make sure what’s been entered into different databases perfectly matches what’s been stored from physical scans.”
These might not be headline-grabbing applications, but they’re high-impact aids to a lab’s daily operations. “A lot of those edge cases come in relatively small but happen frequently,” Richardson said. “When they do happen, it really does slow down the lab. Removing those is probably the best use case for AI right now.”
Both speakers acknowledged that AI-assisted diagnosis for complex cases remains some way off. But the administrative and workflow layer, where accuracy matters but the stakes of an error are less immediately clinical, is ripe for transformation.
Can clinicians trust AI in diagnostics?
Host Peter Birch pushed the conversation toward the human dimension of AI adoption.
The discussion noted that the challenge isn’t purely technical, as clinicians carry personal accountability for diagnostic decisions. Asking them to trust a tool they don’t fully understand highlights the importance of governance and trust.
But Burke also made that point that not using AI is itself a big decision to make. “If you’re not making use of all of the information that’s available out there, that in itself might have risk, because maybe you don’t reach the best conclusion.”
He noted that the generational divide in patient attitudes is already visible: “You can have people who only trust the GP to explain results to them. At the same time, we’re seeing people feeding their pathology results into AI, then going to the doctor saying, ‘AI tells me this, what are you going to do about it?’“
How falling genomics costs are enabling preventive healthcare
Richardson brought the conversation to its most forward-looking point with a striking data arc. Genomic sequencing cost roughly $100 million in the early 2000s. Today it’s under $100. Some projections put it on a path toward less than $10.
“We are getting to a really interesting point,” Richardson said. “We’ve got this acute shortage of pathologists that seems to be getting worse. We’ve got the most powerful AI models available. And we’ve got the cheapest sequencing ever. The confluence of these factors really gives us the foundations to shift from reactive healthcare to preventive, proactive healthcare.”
That shift, from treating illness after the fact to getting ahead of it, is what excites Richardson most. “We are about to enter a stage where you can actually get ahead of many of your health issues, and this is going to be cost-effective for the system.”
What’s next for Magentus and Labflow?
Burke outlined a clear near-term roadmap. LabConductor tackles interoperability first. Then, throughout the coming financial year, modular bolt-ons for anatomical pathology, genomics, and genetics functions, all designed to implement quickly rather than requiring months of project lead time.
“I think this is the kind of moment for pathology to come out of the back room onto the centre stage,” Burke said. “It’s the enabler of precision medicine. The technology enables us to tackle the workforce challenges. And it’s going to be really exciting times ahead.”
His advice for lab leaders was pragmatic: “In pathology, the safest way to move forward is one step at a time, never try to do it all at once. But you can really start to see some value pretty quickly if people take the first steps.”
The landscape is moving faster than strategy cycles can keep up with, but the core pressures of workforce, demand and consumer expectations aren’t going away. The labs that start moving now, even modestly, will be the ones best positioned when the next wave of capability arrives.
Frequently asked questions
What percentage of clinical decisions depend on pathology? Roughly 70% of clinical decisions depend on a pathology test result, making pathology one of the most critical yet least visible parts of the healthcare system.
How many pathology tests does Australia process each year? Australia processes approximately half a billion pathology tests per year across public and private laboratories.
What is incremental modernisation in pathology? Incremental modernisation is an approach to upgrading laboratory systems by layering new capability on top of existing infrastructure rather than replacing legacy LIMS platforms in a single big-bang migration. It allows labs to deliver value faster while reducing risk and cost.
What is LabConductor? LabConductor is Labflow’s integration layer for pathology laboratories. It connects databases, instruments, and file systems without becoming the master system, enabling labs to modernise without disrupting existing workflows or patient identifiers.
How is AI being used in pathology laboratories today? AI is primarily being used in pathology for workflow optimisation and data integrity, including cross-checking reports against clinical history across multiple databases and validating request form data against physical scans. AI-assisted complex diagnosis remains an emerging capability.
How much does genomic sequencing cost today? Genomic sequencing has fallen from roughly $100 million in the early 2000s to under $100 today, with some projections estimating it could drop below $10, creating the economic foundation for preventive and personalised healthcare at scale.
This article is based on Episode 611 of the Talking HealthTech podcast, hosted by Peter Birch. Listen to the full conversation at TalkingHealthTech.com.
