In the recent episode of Digital Pulse by Pharmatica, host Shubhangi Dua, Podcast Producer and B2B Journalist, sat down with Kate O’Reilly, President and Chair, Board of Directors, Healthcare Businesswomen’s Association Dublin Chapter and Healthcare Transformation Partner, Roche. They talk about the future of AI in pharmaceuticals and life sciences, patient engagement, and healthcare transformation.

O’Reilly draws from her experience as a pharmacist, her studies in neuroscience, digital health and healthcare system innovation. She tells Dua that healthcare needs a foundational mindset shift. Particularly, the industry is encouraged to move to patient-first rather than technology-forward. This means keeping technology as a secondary step to ensuring deeper, more structural patient inclusion.

“Patients really need to have a permanent seat at the decision-making table,” O’Reilly stated, alluding to the fact that healthcare innovation should be anchored in the experiences of the people it ultimately serves.

The former pharmacist believes that technology solely will not determine if healthcare innovations are successful. The vision O’Reilly depicts is of patients becoming active participants across the end-to-end care journey, rather than passive recipients.

Why is Patient Inclusion Key to Successful Innovation in Pharma?

“Patient-centricity” is evidently one of the most used terms in healthcare and pharma over the past decade. However, according to O’Reilly, patient inclusion continues to evolve and remains one of the greatest enablers to the development of successful innovation.

Healthcare requires a systemic change that puts patients at the centre of the decision-making process, O’Reilly discusses. By virtue of living with their conditions, they are experts in their own care and the only individuals involved in every single decision made along the end-to-end care journey. They hold critical insights about their care needs that may not be visible to clinicians or researchers. This is spotlighted as the “missing data,” O’Reilly explains, which can change the trajectory of outcomes using technology.

Also Read: Strategic Pharma Intelligence Drives Better Decisions in Life Sciences

How can AI drive more patient-centred healthcare?

Instead of bringing lists of symptoms, patients can now bring clinicians' recommendations, summaries and individualised health insights, with the advent of AI platforms like ChatGPT. However, the real opportunity for AI lies in how patients communicate with their clinical teams, O’Reilly believes.

One example of this is its potential role in shared decision-making, an important component of “patient-centred care”, she explains.

In this context, there is a potential role for AI when it comes to better informing patients and supporting them in taking a more active role in their care journeys, and also for clinicians when it comes to preparing medical information in a format that is more “patient-oriented”. However, AI will not replace clinicians, O’Reilly asserts. AI is an augmentation tool: “It's not a substitute for medical expertise,” she explained.

Also Read: First Quarter 2026 State of Digital Health Funding

How AI is changing patient outcomes in pharma

Healthcare Transformation Partner at Roche illustrates favourable outcomes for patient-first technologies.

One of the greatest values of digital innovation is its capacity for continuous data collection. Conditions like MS often involve symptom fluctuations over time. This means that isolated appointments often only capture a discrete snapshot of the patient's experience at a particular moment in time. Continuous data collection can help create a more accurate and nuanced understanding of patients' well-being.

As excitement for AI grows in healthcare, organisations risk getting distracted by the technology itself. Instead, O’Reilly calls for “needs-led” innovation, which involves rigorous, evidence-based assessment of the problem being solved for.

As she summarised, "We need to fall in love with the problem" in healthcare; creating great solutions means little if the initial problem was misunderstood.

Also Read: Where Healthcare AI Investment Is Going — And Where It Isn't

Key Takeaways

• True patient engagement ensures patients have a permanent seat at the decision-making table
• When it comes to advancing patient-centred healthcare, some of the greatest potential for AI lies in its capacity to transform how clinicians and patients communicate, and how patients engage in decisions about their care
• Healthcare transformation should be underpinned by a “needs-led”, rather than solution-driven approach; otherwise we risk building brilliant solutions for problems that may not exist.
• The user experience of the clinician cannot be ignored. Innovation only moves as fast as clinical adoption. Their readiness to embrace new tools is the bridge to patient impact.

Chapters

• 00:00 Introduction to Healthcare Transformation
• 03:49 The Role of Patient Engagement
• 08:50 AI's Impact on Patient-Provider Communication
• 12:49 Operationalising Patient Engagement
• 14:12 AI and Digital Health Innovations
• 18:31 Collaborative Initiatives in Healthcare
• 20:48 Future of Pharma and Patient Interaction
• 22:46 Integrating Real-World Insights
• 25:26 Patient-Centric Clinical Studies
• 27:28 Key Takeaways for Pharma Leaders

#PharmaAI #MachineLearning #Pharmatica #GenerativeAI #DigitalPulse #RegulatoryAffairs #HealthTech #AIValidation

The main issue, according to Masad, is that the industry confuses two fundamentally different technologies. "Traditional machine learning models are trained to do one specific task very well," she explained, alluding to examples like deviation rates, batch records, and sensor readings. These models are deterministic, explainable, and highly auditable, which makes them well-suited to a GXP environment.

In the recent episode of the Digital Pulse podcast, host Shubhangi Dua, Podcast Producer and B2B Journalist, sat down with Masad to clarify the key distinctions in AI and machine learning models. Furthermore, Masad explains how pharmaceutical companies benefit from using AI models.

LLMs function differently. They are trained on vast amounts of text and generate probabilistic outputs rather than fixed ones. They are designed to be generalists. "The same model that can draft a deviation investigation, for example, can also rewrite a poem," she said. This flexibility is also a challenge: "You can't create a fixed qualification protocol for a model that can respond in a thousand different ways."

Masad points out that this difference has real financial implications. "I have seen organisations default to large language models for everything because that’s what’s in the news, when in reality, a well-trained machine learning model would have been faster to validate, cheaper to maintain, and more defensible in front of an inspector."

Also Read: Where Healthcare AI Investment Is Going — And Where It Isn't

Where AI Is Actually Delivering Value Right Now

When asked where AI is truly making an impact in pharma, Masad provided a practical answer: "more areas than people think, but fewer than what vendors would like to suggest."

In quality assurance, she highlighted inspection readiness and deviation management as key applications. Identifying risk signals from CAPA records and trending data "before it becomes a 483 observation by the US FDA," she stated, is genuinely valuable.

Masad described document processing tasks like gap analysis against changing guidelines, regulatory intelligence monitoring, and drafting response narratives as the current focus. However, she is realistic about the limits:

"Full-scale regulatory submission support is still a few years away from being valid in most markets and for most regulatory bodies."

The use case she finds most promising is inspection risk prediction. She believes it offers "a clear return on investment, a clear validation pathway, and a clear regulatory rationale" because it relies on machine learning instead of generative AI.

Also Read: Reimagine AI-driven Drug Discovery with Pharmaceutical Superintelligence

Where Leaders Should Focus for the Next Decade

Looking ahead, Masad identified three priorities for pharma organisations serious about long-term AI capability, ranked in order: people, regulatory understanding, and localisation.

Regarding people, she argued that the main barrier is not access to technology but having scientists, QA professionals, and leaders who can use it responsibly. On regulatory understanding, she predicted that future leaders "won't just be the ones with the best models" but those who can advance models through validation and approval the quickest.

She also shared insights from her health-tech startup GeneAId Ltd, which applies machine learning to genetic variant classification for underrepresented Gulf and Arab populations. She emphasised that AI designed for US and/or European markets does not automatically apply elsewhere, making localisation a critical blind spot for global pharma companies.

Her final message tied the three priorities together: "We build a foundation, not a series of one-off projects. AI should compound over time, but only if you have created something worth compounding on."

Also Read: What Pharma R&D Tech Investment Committees Actually Fund

Key Takeaways:

1. ML is deterministic and auditable; LLMs are probabilistic and harder to validate.
2. The real AI value today is in deviation management and document processing, not full regulatory submissions.
3. Validation needs scoped use cases and human review gates instead of fixed test sets.
4. "Human in the loop" isn't enough without clear rubrics and continuous monitoring.
5. People, regulatory fluency, and localisation, not models, will determine who wins.

Chapters

• 00:00 Introduction to AI in Pharma
• 03:02 Understanding Machine Learning vs. Large Language Models
• 06:00 AI's Role in Pharma Business Processes
• 09:09 Challenges of AI in Regulated Environments
• 12:11 Practical Applications of AI in Pharma
• 15:12 Balancing Accuracy and Explainability
• 17:55 Responsible Adoption of AI in Pharma
• 20:57 Regulatory Oversight and Human Review
• 23:58 Continuous Monitoring and Validation
• 26:55 Building Long-term AI Capability in Pharma

For more industry-leading insights on AI in life sciences, visit pharmatica.io.

Reach out to Lara Masad here: https://www.linkedin.com/in/laramasad/

Follow Pharmatica.io:

Pharmatica.io YouTube: @Pharmatica_io

Pharmatica.io LinkedIn: https://www.linkedin.com/company/pharmatica-io/?viewAsMember=true

#PharmaAI #MachineLearning #Pharmatica #GenerativeAI #DigitalPulse #RegulatoryAffairs #HealthTech #AIValidation

Judit Mora, CEO and co-founder of Nuumad, joined Trisha Pillay on Digital Pulse to make the case that the route through the NHS is not the only route and that the channel most pharma companies have systematically ignored is precisely where the opportunity sits.

The Blindspot That Is Costing Market Access

Mora opens with a structural diagnosis that should concern any commercial or market access leader. When products are developed, the thinking defaults to two audiences: the patient and the clinician. Pharmacy which dispenses the product most of the time is an afterthought. This is due to innovation funding following the same logic as product development thinking; the gap compounds.

"Because this blind spot exists for big pharma, innovation funding doesn't even flow there," Mora says. "It's a self-perpetuating cycle, a lot of tech innovation comes out from incubators run by big pharma, and it's all linked to expectations on their product pipelines."

The second failure point is equally familiar to those who have watched digital health initiatives stall: patient-facing innovation built without clinical pathways behind it. Mora's example is Babylon Health, a platform that positioned itself as AI-driven, relied on large volumes of healthcare professionals doing manual work behind the interface, and ultimately couldn't scale because real clinical triage doesn't follow a simple decision tree. "When companies launch 'this is a great patient app', what happens when you actually need clinical intervention? It's always an afterthought."

Why the NHS Timeline Is a Strategic Problem, Not Just a Regulatory One

Mora is measured about public healthcare systems. They are not broken. They are stretched, and the consequences of that stretch land directly on patient access timelines. The evaluation process the NHS runs is thorough by design: guideline fit is assessed first, cost is scrutinised after, with multiple steps between authorisation and formulary inclusion. For blockbuster products, the biologics in immunology that represent genuine step-changes in patient outcomes, even those remain second-line treatments not because the clinical evidence is weak, but because they are too expensive to deploy at a population scale. "A new product might take three years to get into the NHS and get in front of patients. If you're ill and there's a product that will change your quality of life, that's a significant burden."

The private route Nuumad operates within doesn't displace the NHS pathway. It runs alongside it. Get an independent medical evaluation, make the product available through pharmacies or private clinics, and let patients who want earlier access make that choice. "It's a much quicker market access, and it opens up the possibility." The equity question is real, and Mora acknowledges it. But availability is a precondition for access of any kind.

The Mechanism: A Prescription Without a Prescription

The model Nuumad has built centres on a Patient Group Direction, the same legal framework that enables NHS flu and COVID vaccination services to be delivered by pharmacy technicians and nurses without individual prescriptions. A PGD defines inclusion and exclusion criteria; a clinician who works through those criteria can dispense the product directly. No GP referral. No prescriber in the chain.

What Nuumad adds is the clinical user experience layer that turns that legal document into a functional digital platform, one that guides a pharmacist or pharmacy technician through a gold-standard consultation, building clinical confidence as it does. "What we want to do is instil process thinking, even for non-prescribing clinicians who may have never run these types of services." The design goal is explicit: the platform should reduce anxiety, not create it. A clinician using it for the first time should feel guided, not exposed.

On AI: Why Nuumad Is Deliberately Not Going There

Mora's position on AI in clinical workflows is a useful corrective to the current market noise. Nuumad uses AI for operational purposes only. Clinical decision support runs on deterministic, rule-based algorithms, and the reasoning is worth understanding. AI outputs vary when models are retrained or updated. In a clinical decision context, that inconsistency is not acceptable. European healthcare data models differ materially from the US datasets most large AI systems are trained on. And critically, AI tools in clinical settings still require human validation, which undermines the efficiency case entirely. "If you outsource your own healthcare thinking to a tool that may not give you the right answer, and in the process is not sustainable, why are you using it?" The accountability point is the one most operators haven't fully absorbed: when an AI tool contributes to a clinical error, the liability sits with the clinician who used it, not the manufacturer. "Very often when I ask pharmacists about this, they say: I didn't know that."

The Commercial Opportunity Pharma Keeps Overlooking

Mora's closing argument is aimed directly at commercial strategy. Pharmacy is an underserved, high-frequency patient touchpoint. The regulatory infrastructure to use it as a market access channel exists. Patient demand is there. What has been missing is the organisational willingness, particularly in larger businesses, to pursue a route that doesn't have the NHS as the primary buyer. "Pharmacy is a big opportunity for businesses that want to innovate and are forward-thinking. Not to ignore pharmacy because it has its struggles, it is definitely ripe for new things." For pharma executives looking at three-year NHS timelines and asking if there is a faster way: there is. For more information on what was spoken about, visit Nuumad or connect with Judit Mora on LinkedIn.

Takeaways

• Market access and new models for therapies.
• The untapped potential of pharmacy in healthcare.
• The role of user experience and product design in healthcare.
• Impact of public healthcare systems on innovation.
• Use of AI and digital tools in healthcare.

Chapters

00:00 Introduction to Market Access in Pharma

03:53 Judith Mora's Journey in Pharma

07:53 Gaps in Pharmaceutical Innovation and Patient Access

11:53 Public Health Systems and Innovation Barriers

16:07 Innovative Models for Non-Prescribing Healthcare Professionals

19:58 User Experience and Product Design in Pharmacy

24:10 AI in Healthcare: Opportunities and Cautions

30:04 Closing Thoughts on Pharmacy Innovation

Adriaan Kruger, founder and CEO of nuvoteQ has spent twelve years building the infrastructure to close that gap. His conversation with Trisha Pillay on Digital Pulse is one of the more candid accounts available of what digital adoption in clinical research actually looks like when the constraints are real.

Why Paper Persists and What It Actually Costs

The persistence of paper in LMIC clinical research settings is not a technology problem. The technology exists.

"Our industry is so hesitant and so resistant to change," he says. "They are so risk-averse — which we totally get in this highly regulated world. These researchers are so worried about potentially using a system and then the data gets lost, or the system is down. So they go back to their tried and tested way of doing things, which is on paper."

The financial dimension compounds the risk-aversion. Funding does not flow to research sites in sufficient volumes to support digital infrastructure investment. The result is double data entry, a risk mitigation practice where patient data recorded on paper is manually typed into a centralised system twice by different operators, then compared for discrepancies. In developed markets, this practice has largely been eliminated. Across much of Africa, it remains the norm.

The downstream cost is significant as data science teams spend extensive time converting inconsistent paper records, Excel files, and incorrectly formatted entries into globally standardised formats before a single submission can be filed.

Real-Time Visibility Changes the Risk Calculation

The operational case for digital infrastructure is sharpest when Kruger describes what real-time data visibility actually enables at the trial level. Clinical trials running across multiple countries like South Africa, Mozambique, Ghana, Finland, and Australia simultaneously have historically operated with delayed information flows. A serious adverse event occurring overnight in South America might take days to trigger a coordinated global response. The pharmaceutical company, the investigators, and the data safety monitoring board are all operating on different information timelines.

"If they don't have real-time data, often there's a delayed reaction on how you deal with things," Kruger says. "Now if they have a real-time view of everything happening globally, they can make insightful decisions, let's pivot, let's change the dosage, or there was a serious adverse event, it happened last night in South America. The morning we wake up, everyone knows."

For pharma executives managing trial costs, trials are expensive precisely because they run long, across many sites, with significant overhead. The ability to make earlier decisions about dosage adjustment, site performance, or trial termination is a direct cost lever. "You can make decisions on whether to move faster, change direction, or potentially stop a study so that you don't spend more money on it."

AI Where It Works in Clinical Research

Kruger's position on AI in clinical research is one of the more technically grounded takes in the current conversation. nuvoteQ is not AI-averse, but the firm is precise about where AI delivers and where the industry's enthusiasm is running ahead of its readiness.

The clearest proof point is a regulatory review tool built for South Africa's medicines regulator, SAHPRA, in partnership with a Seattle-based philanthropic funder. The challenge was a generic drug dossier, submissions that can run to 15,000 pages, were sitting in regulatory queues for up to nine years across Africa, delaying the availability of affordable medicines for the patients who need them most.

The tech company deployed an open-source LLM locally within SAHPRA's data centre, with no external internet access, trained to analyse dossiers and produce a two-page risk summary with hyperlinked references to the source document. "That has reduced the timeline to review from where it typically takes five to six weeks. We can do it in about 45 minutes now."

The model works because the data never leaves a closed environment, the task is well-defined and repetitive, and human reviewers remain in the loop for final decisions. That last point is deliberate. "You will always have a human in the loop. There will always be a healthcare professional. AI will help do some of the number crunching, but ultimately, when it comes to patient care, the ethical component is how patients are being treated. I don't think AI is going to be there for a long time."

The liability question Kruger raises for clinical AI tools mirrors what is emerging elsewhere in regulated industries: the responsibility for an AI-assisted decision sits with the clinician or researcher who acted on it, not the software provider.

Africa as a Strategic Asset

Kruger closes with a reframe that deserves attention from anyone structuring a global trial portfolio. South Africa's clinical research infrastructure, built over decades around institutions like the University of the Free State in Bloemfontein, is world-class, not despite its context but in part because of it.

"Because of our diverse population just in our country, we are world leaders when it comes to clinical research. During the COVID pandemic, the fact that we discovered some of those strains that's not by chance. It's because we are world-class."

Africa's genetic diversity is, he argues, an asset that the global pharmaceutical industry has consistently undervalued. As precision medicine and genomic-driven treatment personalisation move from oncology into broader therapeutic areas, that diversity becomes a competitive advantage for trial design and drug development that no other region can replicate.

The infrastructure to unlock it is digital, connected, and validated to FDA and EMA standards, which is what Kruger is building. The industry's willingness to fund and support it is what will determine how quickly African patients stop waiting.

Learn more about nuvoteQ and connect with Adriaan Kruger on LinkedIn.

Takeaways

• Digital transformation in clinical research
• Barriers to digital adoption in pharma
• Role of AI in clinical trials
• Future trends in personalised medicine
• Data sovereignty and privacy in healthcare

Chapters

00:00 Transforming Clinical Research: The Digital Shift

03:01 Barriers to Digital Adoption in Pharma

08:13 Data Accuracy and Double Data Entry Risks

11:07 The Role of AI in Clinical Research

19:21 Real-Time Data Visibility in Trials

23:44 Scaling Tech Solutions in Africa

28:36 The Future of Clinical Research: A Vision for 5 Years