Harnessing Transformative Conversational AI to unlock Highly Improved Engagements

We are pleased to welcome Thomas Mrosk and Manuel Mitola, Co-Founders of ctcHealth. Drawing on their deep experience in applying AI to real-world healthcare challenges, they share how digital tools can empower clinicians, field teams, and patients alike.

Karsten: Thank you for taking the time for this interview, Thomas and Manuel. To get started, what motivated you to found ctcHealth, and how has your vision evolved since the early days?

Thomas: Thank you for having us. The moment for me was really November 2022, the ChatGPT launch. When I first experienced what this technology could do, I was immediately fascinated. I just thought: this is going to completely change how we do pharmaceutical marketing and sales. It wasn’t about tweaking existing processes; it was about reimagining them from scratch. That was the moment I knew I didn’t just want to observe this shift. I wanted to be part of it, to help shape it. And I saw massive opportunity in doing that.

The first use case that made total sense was training. Traditional training, whether it’s reading, videos, or e-learning, is always passive. With conversational AI, we suddenly had a way to make learning active. People can simulate real situations, apply what they’ve learned, get feedback in real-time. That’s a fundamental shift.

We started collaborating with subject matter experts and HCPs from the very beginning. Working with them revealed just how much potential this technology holds. The same tech we use in training now applies to medical education, and even more exciting, patient support. Imagine having a 24/7 tutor for patients, someone who’s knowledgeable, empathic, and understands the patient journey better than anyone. That’s where this is going.

Funny enough, when we started in early 2023, we thought we might already be a bit late. But the speed at which the technology has evolved since then, it’s been wild. In just 18 months, what’s become possible is beyond anything we imagined at the start. And it’s only accelerating.

Karsten: AI is gaining traction across healthcare. Where do you see the biggest untapped opportunities today, and where do you advise more caution?

Manuel: Honestly, it’s hard to think of any area that won’t be impacted by this transformation. If we look at pharmaceutical marketing, for example, one of the big dreams of any marketer, hyper-personalized customer communication, is suddenly within reach. With generative AI, we’ll be able to use CRM data and other sources to tailor messages in real time, automate large parts of the MLR review, and even identify the best channel for each customer - all done dynamically. What used to take months from insight to content to delivery will soon happen in days, maybe even hours. That fundamentally changes how marketing works and what marketers do.

And that’s the broader point: no matter where you work today, it’s not so much what you do that will change, but how you do it. Generative AI shifts the whole approach.

But there are areas where caution is needed. In the short term, some things may be overhyped, but in the mid to long term, it’s almost certainly underhyped. The speed of change is incredible. Whatever setup you’re building today might be outdated six months from now. So, flexibility is key.

Karsten: From your perspective, what are the main barriers to adoption when it comes to AI-based tools within pharma and healthcare environments?

Thomas: One of the biggest barriers we’re seeing right now is lack of in depth understanding and education around the technology. We’re coaching a lot of senior leaders across pharma, but the need for updated, high-quality and industry specific information is still high.

There’s also a clear generational divide. I recently heard Sam Altman talk about this, and it really stuck with me. He said they see three groups among ChatGPT users: Gen Alpha, who treat it like an operating system, they’re building agents and embedding it in their workflows already. Then there’s the 25–35 crowd, using it as a life advisor. And above that, most people still use it like Google, a search engine. And that gap in how people use it really matters.

This technology is incredibly powerful. But if you don’t understand what it can do or how to apply it, your organization won’t move, and you risk wasting time and money. Right now, it all comes down to AI literacy. And that’s not something you pick up in a single training session. It’s like learning a language. It takes ongoing learning, upskilling, and embedding that knowledge across teams, across generations, and across functions.

Karsten: Looking beyond the buzz: What makes an AI solution actually useful in day-to-day clinical or commercial settings?

Thomas: There are so many useful AI use cases we could talk about, but if you zoom out a bit, I think there are two big ways organizations can benefit from AI right now.

First, and maybe the easiest entry point, is productivity. Especially when it comes to knowledge work. I recently read that the average office worker spends 32 days a year just looking for information. And it checks out…“Where was that SOP again? What exactly did it say? What was the approved claim here?” These are the kinds of daily time drains we all know. With AI, you can flip that. Instead of searching, you ask, and get an answer. With in-house available tools like Microsoft Copilot Studio, it's now totally possible to set up AI agents that pull from your internal systems and give you what you need instantly. The time savings can be huge.

But the part I find much more exciting is what happens when you go beyond productivity and start tapping into harnessing intelligence of these models. Mo Gawdat recently made a point that really stuck with me:  using ChatGPT is like gaining 25 IQ points. These models already operate in many domains at a level comparable to an IQ of 155. Einstein is estimated to having had 160. So, imagine what happens when you bring that kind of additional intelligence into your daily decision-making.

That’s what we’re doing a lot for ourselves, building AI agents that act as domain experts and thought partners. Whether it’s a medical strategy advisor, a competitive landscape guide, or a content quality reviewer, these agents can help you think better, not just work faster. And while we’re just at the beginning of integrating that intelligence into corporate life, it’s where the real value lies in my view.

Karsten: How do you see the role of Medical Affairs changing in the next 3–5 years as digital and AI capabilities mature?

Manuel: Just like in marketing, there’s massive potential for acceleration. One of the core responsibilities of medical affairs is identifying evidence gaps and closing them quickly. With AI, insight generation, from HCPs, field teams, or advisors, can happen much earlier and more efficiently. That means data generation and publications can move faster, with more precision and impact. It’s a real shift from reactive to proactive.

Another exciting area is smarter clinical trial design. Imagine having AI agents trained on your entire body of internal evidence, competitor data, and broader scientific literature, designed to think like top experts in the field. These agents can serve as thought partners, helping teams brainstorm trial designs that are more impactful, cost-effective, and quicker to execute. This kind of strategic support can dramatically reshape how trials are conceptualized and run.

At the same time, medical affairs needs to prepare for how physicians themselves will start embedding AI into daily practice. Healthcare is one of the most promising domains for AI. We’re already seeing models that outperform doctors in diagnostics and pattern recognition, especially in areas like radiology. These tools won’t just boost productivity, they’ll enhance patient outcomes. While adoption may take time due to the more “traditional nature” of healthcare systems, the transformation is inevitable.

All of this points to a broader evolution: medical affairs is becoming even more central to strategy. It will be about leading, not just supporting, using AI to drive smarter, faster, and more meaningful decisions across the lifecycle.

Karsten: In your experience, what distinguishes pharma organisations that succeed in embedding innovation from those that don’t move past the pilot phase?

Manuel: The first, and probably most underestimated, factor is AI literacy across all levels of the organization. It really starts with senior and mid-level leaders. They need to understand not just the potential of the technology, but also how to spot use cases that are actually relevant to their functions. It sounds simple, but in practice, this is where many efforts stall.

Once there’s that basic understanding, the next step is to focus on quick wins, use cases that are small, manageable, and deliver value fast. What we often see instead are overly ambitious plans, like rolling out a GenAI training solution globally in one go. It sounds visionary, but the reality is: it doesn’t work. Different regulations, market needs, and stakeholder layers quickly become roadblocks. The better approach is bottom-up, start small, learn, and scale responsibly.

Then there’s the operating model. GenAI doesn’t fit neatly into the systems pharma companies already have. Approval processes, quality checks, governance, they all need to evolve. We’re dealing with non-deterministic systems now. You need new ways to validate the reliability and accuracy of outputs, and new frameworks for reviewing AI-generated content. Most companies haven’t figured that out yet.

And finally, speed. The tech is moving incredibly fast. Many organizations set up global governance boards for AI, centralized committees that quickly become overwhelmed and end up blocking progress. Governance is critical, yes, but it needs to be smart and agile. If you don’t fix that, you’ll never get beyond the pilot.

And to be very honest, we haven’t worked with a single Big Pharma company yet that’s truly done all this well. Most are still stuck in pilot mode, not because the tech isn’t ready, but because the organization isn’t.

Illustration: Embedding Conversational AI in Pharma

Karsten: A big thank you to Thomas and Manuel for this insightful conversation. Your reflections made clear that conversational AI is no longer just a tool for boosting productivity, it’s becoming a true thought partner, capable of enhancing learning, decision-making, and patient support across healthcare.

To succeed, pharma organisations must build AI literacy at all levels and adopt a flexible, fast-learning approach to implementation. As AI capabilities mature, Medical Affairs is set to take on a far more strategic role, accelerating evidence generation, shaping smarter trials, and leading the shift toward more proactive, data-driven engagement. #ConversationalAI #PharmaInnovation #AIInHealthcare