Medidata NEXT New York 2026 LIVE

NEXT New York is our flagship event, bringing together the world’s leading life sciences companies and thinkers to envision a transformative future for healthcare that challenges our preconceptions of what’s possible in clinical trials.

This special edition of from Dreamers to Disruptors was recorded live from the show floor at NEXT, capturing the hottest takes and insights from Medidata leaders and special guests. From the end-to-end AI support reshaping studies, to building trust in cutting-edge clinical technologies, our speakers share their fresh perspectives on what matters most to pharma professionals today.

Session 1 Speakers:

• Edward Ford | Vice President, Client Engagement, Medidata
• Anthony Costello | CEO, Medidata
• Matteo di Tommaso | Senior Vice President, R&D Business Insights & Technology, Bristol Myers Squibb
• Nimita Limaye | Research Vice President, Life Sciences R&D Strategy and Technology, IDC

Session 2 Speakers:

• Aryana Hosseinkhani | Vice President, Head of Product Marketing, Medidata
• Meghan Harrington | Vice President, Clinical Trial Financial Management, Medidata
• Paul O'Donohoe | Senior Director, eCOA Product and Science, Medidata
• Sharlene Carnegie | Senior Vice President, Product and Engineering, Medidata

Session 3 Speakers:

• Jeff Ventimiglia | Senior Vice President, Offering and Portfolio Management, Medidata
• Bryant Fields | Vice President, Partnerships, Medidata
• Robert Lyons | Vice President, Engineering, Medidata
• Jia Chen | Head, Product, AI, Medidata

Session 4 Speakers:

• Lisa Moneymaker | Senior Vice President, Chief Strategy Officer, Medidata
• Tom Doyle | Senior Vice President, Chief Technology Officer, Medidata
• Claire Biot | Vice President, Life Sciences and Healthcare Industry, Dassault Systèmes
• Patrick Johnson | EVP, Corporate Research and Sciences, Dassault Systèmes

Visit our website

Read our blog

Follow Medidata: LinkedIn, Instagram, X, YouTube

Edward Ford: All right, we are live. So in 1999 some founders put a computer in a rolling chair, rolled it down the street, and that was the inception of Medidata.

Here we sit, almost 27 years later, kicking off next New York live, from Times Square.

Welcome to from Dreamers to Disruptors. Live. This is our first live podcast. Go ahead and make some noise.

You hear that live studio audience in the background. I'm so excited to kick this off.

My name is Edward Ford. I have an illustrious panel of guests with me to my right from Bristol, Myers, Squibb, Matteo di Tommaso.

We have IDC analyst Namita LeMay, and we, of course, are accompanied by a from Dreamers to Disruptors, a mainstay Medidata CEO Anthony Costello, welcome to you all.

Thank you so much for taking some time to be here with us today. Make some noise for the panel.

Y'all, yes, yes, yes. So we just stepped off, literally just stepped off of the keynote stage, and the excitement in this room is bubbling so quick reactions.

Matteo, I'll start with you. What did you hear from the keynote stage that got you excited or made you think a little bit?

Matteo di Tommaso: I think the biggest thing I saw Edward that made me excited was just the span of things that Medidata is tackling in parallel, and the very good patient focus on the way you're looking at things.

It's all about making sure we make a difference for patients in the end of the day.

Edward Ford: Absolutely. Thank you. Namita, over to you. I know you.

You see a lot of industry conferences, did anything stand out for you today in that opening session?

Nimita Limaye: Yeah. I mean, I was just telling Anthony I thought the trial simulations piece was, was fantastic.

Yeah. I think you spoke about probability of success, and I really feel that that is going to make a difference.

Edward Ford: Nice, nice. Now, Anthony, obviously you were on stage orchestrating the entire keynote.

You were like of, the opening session over there. You know, what are some of the things that stood out to you in the run up to today and that you were really excited to share with the broader Medidata community.

Anthony Costello: Yeah. Well, you know, today is a big day for us, because next New York is always big.

But the run up to next New York really has been going on for a while, because we used to do next New York in the late fall, in November last year, we decided to move it to the spring.

So we had, we kind of skipped a year, and we move now into doing our whole next series in year, starting in March in New York, and then all the rest of the cities around the world for the rest of 2026 so it's been a pretty big run up.

And I think one of the reasons the keynote today was so exciting for me is that we really had all of 2025 Medidata advancements getting ready for this next 2026, and I hope it, I hope it really showed,

Edward Ford: yeah, I think it came through. I mean, obviously, working with you to deliver all this, and seeing the energy in the teams to deliver this, and as they, as they make those advancements, sharing those things out, I think you definitely express that from the stage.

And I think it's taking us in a great direction.

Namita, I want to ask you about kind of where we are relative to kind of AI adoption across the industry.

I mean, there's a lot of a range of talk. Everyone is obviously talking about AI.

You can't help but not. I'm curious about your perspective on where the industry is relative to adopting the latest in some of the generative AI capabilities, and what that value is looking like in terms of transforming some of the overall outcomes that our industry sees.

Nimita Limaye: Sure, as you said, AI is disrupting everything.

I think it's an apt title for this panel, because it's disruption at scale.

It's so much that's happening in the world right now, but AI is a major disruptor.

I feel we have gone through various phases. So, for example, in 23 we were going through what I would call the pilot-itis phase, right?

I mean, every everything was being tested out. There was this huge FOMO, and everyone was like, Okay, we got to do something, not sure what, but let's try it out.

A 24/25 I think governance is set in. Platformization is coming in. There is, there is a clear vision that we need to embed this within workflows.

So those are the perspectives that are changing. Just a clinical trial optimization.

You spoke a lot about that at IDC. We did a survey right? And we saw that.

We did the survey late last year, and we saw that about 34% of the industry was still doing pilots, but about 38% Sent said, Hey, we are already implementing this.

We're already rolling it out two years down the line.

I think what you're going to see is deployment at scale.

So there is an urgency, there is acceptance, trust is definitely a challenge. Change management is a critical component.

But will this scale? It will great to see the initiatives that you're taking from the age identification, part of it, because it's Gen AI and Agent AI, which are going to takes two to tango, and they're going to work hand in hand.

So I think this is the future

Edward Ford: that's that's amazing. And, you know, you said two things, and actually one thing, I want to give Matteo a little bit of time to talk a little bit more about.

You talked about this idea of driving adoption and how you do that at scale within an organization.

Matteo, you started talking about this a little bit on the on the keynote, and how you look to drive adoption within your organization.

I'm just curious to hear more about kind of your philosophy around driving organizational change specifically targeted at AI, because the reality is, people don't like to change, even with the great promise of this and even as we pursue this mission.

We talked about serving patients, when folks now get into their workflows, they have a workflow, and yet, maybe 62 steps that are held together with duct tape and bubble gum.

But that's their process, and it can be difficult to get them off their process.

So I'm curious how you go about addressing that and inspiring the kind of catalyzing the desire within the organization to change and adopt some of these new ways of working.

Matteo di Tommaso: Well, I've got to start though by saying that I'm glad to hear from Namita that there is a possibility that these things will scale, because I now have aI having an impact across the whole value chain of R and D, and we have three platforms now that are, that are, are truly scaled, that we have 320 some clinical trials, all using our development operations platform that's AI powered.

We have the ability to use AI to work with our portfolio data, and we have a document authoring platform that helps us to accelerate document creation all the way for study startup, all the way through submission.

And so it is scaling today is certainly where I see it.

And how do you how do you do that? It's not the the way you do that is a couple of things.

But number one is getting the right people on the team from the beginning, making sure the mission of the team is really crisp and clear, and making sure that you're paying attention to execution, execution, execution, every quarter and every step, and not just to build platforms or have AI, but to make a real difference to medicines, ultimately.

And so it's the impact, measurable impact, on our work to bring medicines to patients, that's the real driver, and that it's easy to be motivated about that and excited about that.

I guess it's table stakes, Edward, that you got to have something that's worth using, right?

You got to have something that really is game changing or really makes a difference.

Edward Ford: Yeah, the trust there, I think you talked about, within those teams of lining up with the mission of serving patients is critical.

Having tools that actually do what they're promising to do also helps to reinforce and give people a little comfort about moving away from maybe something that they were used to using and dipping a toe into something that might be a little bit different.

Anthony, this concept of trust, right? Balancing this off, because you said it a couple times, you Tom as well, talking about AI technology, and specifically on the Medidata platform, not being this quote, unquote black box where it some compute happens, and then you spit something out, and being very, very keen on, hey, everything that we're doing within the Medidata platform relative to AI, it's traceable.

Obviously there's, there's regulatory implications across across the platform, for that talk a little bit about building trust in this era of AI, for folks that are using the tools, but also for the output, talking about regulators, talking about the industry and really getting back to patients feeling like they can actually trust these outputs.

Anthony Costello: Yeah, yeah. Well, I think you know, one of the reasons we've invested so heavily in this area over the last few years is we view ourselves as nicely positioned at the center of so many different clinical trials where we already have trust.

We already have huge portions of our customers portfolios run in some part, or in some cases, in every in every way, on the Medidata platform.

And you know, I think everyone, we haven't said it yet on the podcast, but everyone here knows that a big part of moving from.

What Namita referred to as the FOMO era, or the pilot era of AI, into something that can be done at scale.

A big part of that is having AI that's useful to you.

It's become such a buzzword that there's junk all over the place, junk AI learning on junk training data sets, and not really bringing a big opportunity or a big advantage, let alone the trust piece, but, but we're different from that, because we've got these 10s of 1000s of historical trials already cleaned, already delivered to regulator regulators or in the process of still being run, live out with sites.

Our AI is training on those data sets with those customers that already use us.

So, you know, that's a big piece. I think the other thing Matteo touched on it for a second, we we've invested heavily in patients.

And I know our whole industry, obviously, is driven around bringing better therapies, faster to patients.

Specifically at Medidata, we built again through Alicia, Alicia Staley's work, we built a patient insights group.

We have a whole cadre of patients that are on staff at Medidata helping us design, not just the software, but the AI, what's going to be useful, what's going to be trusted, and how do we take it out of R and D bubble into real sites, real patients, real trials, in a way that everyone feels comfortable using.

So, you know, we've tried to lean into that trust factor so that we could hopefully have that kind of relationship with customers when these new technologies come out, and trust is really required to get the uptake on those new technologies.

Edward Ford: One thing that has not been spoken at this desk, which I'm actually really excited about, is no one's really mentioned the product name, which is which is pretty cool, and it's part of our philosophy relative to rolling out these three experiences, the study, the patient, the data, experience.

In our next general session, we're going to do a deep dive into each of the three experiences and some of the developments that we've made there.

And so Matteo and maybe Namita, you can respond to this as well.

I'm curious how, because one of the things you said earlier, Mateo, is that you are using AI across not just in one area of your of your of the kind of research cycle, but really across the entirety of that cycle.

So I'm curious in your response to us addressing and thinking about things in a more holistic fashion.

So this idea of, how do we optimize for patient involvement in the study?

How we optimize data collection and analyzation, and not just focusing in on some of these discrete quote, unquote products or narrowing in on a process, is that kind of how you're talking about things and looking at it within your organization?

Matteo di Tommaso: Absolutely, there's no doubt that we've over we have an overly complicated environment in terms of switching between experiences from one place to another.

A big part of the work that we did on development operations was to try to make that a much more streamlined, simpler, single place to go that does guide you from one end to the other.

Now that's always in competition and has the challenge of and I need to be able to swap out modules that do really cool stuff very rapidly.

So how do I maintain that end to end experience that is easy and familiar and comfortable, while still being able to swap out where I'm getting those components, whether I'm building some components in a way that it gives an overall end to end ability to run a clinical trial discover a medicine.

So that's not that's the balancing act we have to play.

And agility is everything, right now, right the mode is gone on.

You can create millions of lines of code with no problem.

So you've got to compete on your ability to really understand how to solve some important problems and really understand your customer base and their experience.

And you got to compete on agility

Edward Ford: and Namita.

I wanted you to continue on that same thought with an industry perspective.

There are many point solutions that do specific pieces of the of this cycle.

And how are you kind of seeing, or are you seeing a shift away from, we do this thing really well to then maybe talking more broadly and holistically about supporting the process of clinical research.

How is that happening in the market, and how are folks talking about that? And how are, and how are also, how are sponsors, and then CROs engaging with providers to support their execution for that?

I'm curious in your perspective, there?

Nimita Limaye: Firstly, I think this is the rev unified platforms.

Everyone is sick of single point solutions and siloed, fragmented systems. There is a five fundamental need to integrate and to drive the connectivity.

And also, I think it's really great that you're weaving in the patient's voice into the products that you're developing, because at the end of the day, it's about the patients.

I think patient experiences everything. We are all patients over here at different points of time.

The other part, I would say, is, and I'll share a small example here.

This was the CFO of a big pharma company who is working with a tech company investing in major, major investments in platforms.

And he said, you know, to your point, you can code everything.

He said, We are building in we are coding ourselves or using agents, and we want to completely disinvest in this platform.

Do you think it's a good idea? And it may be a good idea, but the fundamental question is, what goes into building that?

A lot of domain expertise, a lot of regulatory compliance, and at the end of the day, the data that goes in. I think Hugh called out the piece around the data, and I think everyone's using AI right again, if you just go back to some IDC data, there was a very interesting statistic, which said that 54% of the industry said that using Gen AI for clinical trial optimization is absolutely critical to disrupting the way we operate.

That's huge, right? But Gen AI models may be great, but if you have models that are trained on domain specific data, I think it's it makes a world of difference.

Bringing it all together, weaving in the patient experience and providing a unified platform and models that are really trained on domain specific data.

I think those will be your differentiators.

Edward Ford: All right, so we have a tradition on from Dreamers to Disruptors.

There's a question that we asked to close out every episode.

It's usually Anthony and one other person, but there's, there's, I want three of you.

I want to make sure that you each have a chance to answer this question.

All right, so our question is, What do you dream to disrupt next?

What do you dream today that you can disrupt in the future?

So Anthony, I'm going to give Matteo and Namita just a second to think about it.

Anthony Costello: You know, I think lots of things, but since we only have two and a half minutes left, yeah, I think what I would narrow this down to, and let's go back to the keynote theme around get after it.

So the way we like to end the podcast is, what's your dream, but also have it be a call to action to the audience that's listening, and I think I would follow through with this, get after a call to action to answer this question, we we have the tools, we have the technology.

We're through pilots and FOMO, and we're into really actualizing some of the longtime potential benefit of AI to do something different now.

So I think my call to action for our industry would be, let's do it. Let's get after it.

Let's go faster than is comfortable and really try to make a big difference in the way our industry works over the next few years.

Nimita Limaye: I'd say move away from from anything that's document, move entirely to data, because data is really I started my career, gosh, 30 plus years ago, and as I walked through the clinical trial industry, I was why are we looking at anything beyond data, because it's data that is really the intelligence behind everything.

And today, with AI, no, AI is going to succeed without the data, but we still have a lot of static documents and everything should move into a dynamic, data-driven world.

That's what I believe is the future.

Matteo di Tommaso: And I'm going to go simple, more medicines for more patients faster. And I think there's a lot of promise in the current technologies that we have in the AI to both make that a reality of increasing probability of technical success and probability of success on clinical trials, and to make them go faster so that we get more medicines to more patients.

Edward Ford: Well, Said. Matteo. Namita Anthony, thank you so much for christening our live from Dreamers to Disruptors.

Was it good? Y'all, yes, the crowd is live in here. They're making some noise.

All right, that's it for our first session. We have some more live sessions coming here from next.

We've got a ton more. Got more on the main stage. We've got some breakouts starting this afternoon.

A lot of excitement here. Thank you for joining us, and we'll see you later here live from next, New York.

Make some noise, everybody. You.

Aryana Hosseinkhani: Hi everyone. Welcome to our live broadcasting of from Dreamers to Disruptors from Medidata. Next in New York City, I'm so excited.

We just came off the heels of a beautiful experiences keynote.

I'm joined today by our in charge of the study experience, Megan Harrington, she's bringing the expertise for study.

We've got Paul O’Donohoe, he's bringing the expertise for the Patient Experience.

And then Sharlene Carnegie, she's bringing the expertise for the Data Experience, along with platform.

So you guys, thank you so much for joining today. I'm so excited.

I know there's a ton of like excitement around AI.

Well, let's just focus on the Experiences for a quick minute, and then we'll get into AI.

We know this isn't linear, right? The Experiences, they can start with the patient, then you have the right data sources coming in from the patient, but then how do you derive insights from those data sources to then design the right studies for those patients?

So again, not linear, but let's just start at one starting point.

And I think instead of focusing on execution, like we always do, let's move a little to the left and see how do we design the right study?

How do we use the right insights to design that study?

So Megan, why don't you walk us through what you guys are doing?

Meghan Harrington: Yeah, it's a really exciting time.

We are taking that position of shifting left to execution.

Everyone is focused, rightly so on study startup and site activation and enrollment timelines and really hitting that first patient in and we strongly believe that in order to execute seamlessly, you really need to go left of that and start planning and designing your trials well in advance, perhaps even before you have that concept sheet for a protocol, or if you have a draft protocol, really looking at the insights and what you can do to optimize your operational plan, your financial plan, so that you're setting your teams up for success when they're running those studies.

Earlier today, we heard the stat that one out of 10 studies is successful.

We're wanting to move that needle tomorrow. Four out of five studies are successful, and the way that you can do that is simulating early learning from the data, looking holistically across you know, what is this impact from a site and patient burden perspective?

What will this cost me? What does this look like in terms of the sites I'm going to be able to recruit, getting all of those things in order before you're going into that first site and bringing that first patient to that visit, that's where we feel like, that's where we're investing our time.

That's where we feel like our customers are going to get a lot of value from setting themselves up for execution.

Aryana Hosseinkhani: Thanks, Megan. And so back to not being linear, but we are moving a little linear.

Paul, I'm going to take it to you next on the patient side. Megan just mentioned alleviating patient burden.

And so let's start with the study design really quickly.

What are we doing within the patient experience to enhance maybe eCOA builds making it easier for patients to participate.

How are you seeing the patient experience team really enhancing that patient experience?

Paul O'Donohoe: Yeah, I think it's interesting. Actually, we're almost a victim of our own success when it comes to these technologies.

We heard some really interesting stats about the huge increase in the amount of data points we're capturing in these studies, and specifically when it comes to capturing patient data, it's very easy just to I'll just add in another questionnaire.

I'll just add in another sensor, because it's so easy in these unified platforms to turn on and activate these additional data streams.

But what we really find is that doesn't necessarily bring value ultimately, to your study.

And so spending the time up front to really optimize your study design and really hone in on those key end points, that's where time is well spent to get a really focused endpoint strategy, and then regards to actually capturing the data from the patient, I think clinical research is unique in regards to the technology world, where, you know, in other technologies, it's all about time engaged with the platform, and how long people actually spend using the technologies.

We want to make our technology engaging, but we also want patients to be able to get on with their lives.

And so really, we want to be able to get in and out as quickly as possible, complete the tasks they need to complete surface the next action they need to do, right there in front of them in the my Medidata app, and then they can just get on with their lives.

They're already busy enough managing their disease. We want to make it as easy as possible for patients to complete and provide these these data points.

So that's really what we're focused on within the my Medidata app is allowing us to centralize and surface all that information the patient needs to complete their study activities and get on with their normal lives.

Meghan Harrington: Just to hop in there. And I think Paul one of the kind of other factors is, when you're making it easier for the patient, when you're designing with a patient centric mindset, you're also benefiting the site, making it.

More simple for the patient to interact with the application or the system is going to alleviate the burden from the site from oftentimes they're playing a pseudo Help Desk, or they're asked to be a pseudo tax expert or a pseudo travel agent, if we can get it right for the actual end user, the patient or the CRA or whomever, it has a ripple effect across all of the stakeholders involved in running that clinical trial.

So that's what I'm really excited about.

Paul O'Donohoe: Yeah, it almost feels like historically, there's been kind of a well of burden, and we've just kind of shifted it around to different stakeholders, and it's been on patients.

At times. We're dumping it on sites, often to be that kind of technical support, but really using these technologies to actually meaningfully reduce that burden across all the stakeholders.

Aryana Hosseinkhani: Yep, you guys said some nuggets I want to come back to, specifically around notifications. There's some announcements that were made today travel. There were some announcements made today that I want to come back to.

But you said platform. I noticed Sharlene looked up. So if you ever Sharlene is our expert for platform and data. Paul eloquently said that things are getting more complex. Today on stage, they said things are getting more sophisticated.

So I'm actually going to go ahead with the word sophisticated. But we all know there's a lot more endpoints. There's a lot more complexity and sophistication now to protocols.

How do we actually aggregate all that data, so many data sources coming in? What are we doing the Data Experience, to be able to aggregate that and then be able to derive insights very fast, not to manually go in and see what does this data mean for my study?

How are you guys doing that in the Data Experience?

Sharlene Carnegie: Yeah, that's a great question. Obviously, the landscape is changing at a rapid clip, right?

I think the metric that they called out in the keynote was billions of data points that run through the platform at any given point in time.

So that's an incredible number that is scale. So it's really important to make sure that the foundations are in place such that we get it right.

Foundations of a data platform are so critical to enabling execution, enabling patient interaction.

So it's things like making sure that the data always maintains that consistency, making sure that the data always has that quality.

And it's really building the technological foundations up front such that we can leverage that data downstream and throughout the clinical research life cycle.

One thing that I do want to call out, because it was called out in the keynote, in the experiences keynote, is, you know, as you mentioned, the patient burden.

A patient is already going through so much stress going through the clinical trial itself, and we want to make sure that first point of entry of data from a patient into the my Medidata platform doesn't have to be something that's repeated by them, because we fail to capture that data accurately, or we fail to make that data be made available downstream.

So it's really critical to ensure that we have a solid, reliable, scalable platform at the core of our platform, which, by the way, fun fact, we actually do in Medidata, because we want to reduce the burden on the patient.

We want to reduce the burden on the site such that we can leverage the data for downstream activities, like aI driven insights, etc.

Paul O'Donohoe: I think that's that's consistently what we hear from patients, but also sites around the frustration of why you asked me for this again, I already answered this questionnaire.

I already submitted this data point during so being able to kind of smooth over those and connect up those pieces, really, I think, has a big impact on that burden question again.

Meghan Harrington: I think we're seeing that. I know we heard about this during the Experience keynote. So pre populating the study feasibility or the site feasibility questionnaire, pre setting up your TMF and pre populating some of those documents.

I think one of the comments was, no one went to medical school to write documents or file documents.

Let's let the AI tooling do those kind of simple tasks for the customers so that they can get to the real science, get to the real kind of important human in the loop to make a judgment call on quality or risk

Paul O'Donohoe: and then for the patient perspective, you know that a lot of that data that we need in the clinical trials is already in their health record, so being able to pull that in using health record connect again, rather than just asking the patient to complete a questionnaire that they've already completed maybe a couple of weeks ago.

Sharlene Carnegie: You know, Pascal Daloz, in today's keynote, brought up a very, good point.

And I think that's been a recurrent theme throughout the day, AI is only good as the knowledge or the data that it's grounded against.

So, you know, a lot of times we at Medidata really focus on the concept of data centricity, because without all of that knowledge baked into our systems, without that data, it's really impossible to do things like automatically generate those documents, or pre populate a lot of those things, or identify those examples where there's a conflict and a visit, and therefore it needs to make a decision on how to reschedule without the core data being AI ready.

It's. Impossible to do any of those things,

Aryana Hosseinkhani: you guys. I'll just hit on some things. But for those that dialed in and maybe missed the keynote, let me just level set a little bit.

There was a lot of conversations around AI and the ability to bring in AI, to automate things and reuse data that we naturally already have in our systems, making it more accessible and reusable.

So I really want to hone in on that word reusable you mentioned the feasibility questionnaires.

These are things that sites have to sit there and do instead of being with patients, they have to go back and refill out forms that say, Yes, I do have a minus 20 freezer, which was a comment that was made on stage.

Taking a look at shifting a little bit towards AI, because that is the buzzword, everyone listening is probably very focused on AI. And what does aI mean to the clinical trial space?

Megan, I want to speak to you a little bit. How are we leveraging AI within the study experience?

Meghan Harrington: Really, it's a charge of for our teams to leverage it everywhere we can, where it makes sense, right?

And I think that's actually an important conversation to you. Where does it make sense? Where is it going to provide the greatest value, and not just a distraction or kind of doing it for doing its sake?

So from a product perspective, we are absolutely looking at how we can bring together our already our AI capabilities in simulating enrollment forecasts, simulating operational plans and budgeting.

You know, we are bringing it into looking at monitoring visits and monitoring schedules and pre populating documents and TMF or for site feasibility.

And we're also looking at it in terms of how you financially forecast those particular studies.

I think it's also important to note that as teams internally within R and D, we're using it to change how we work so that we can get work out faster, right?

So we're using AI tooling in the way that we write code, or think about a business requirement from a customer request.

How can we move faster to actually bring value, you know, sooner, to our customers.

Aryana Hosseinkhani: I love that you said that we don't just walk the walk or talk the talk, whatever the saying is, we actually utilize AI ourselves, too.

And so notifications. I mean, we all are pretty addicted to our phones. You get an email, you get a notification, you get a notification to be here right now to record this probably all had calendar holds

AI is driving notifications tailored to patients to make it easier for them to engage with their trial. Maybe someone's on a morning person, maybe someone's a night person, maybe someone operates better on weekends.

Paul, what are we doing with AI and the patient experience and maybe notifications tailored to patients?

Paul O'Donohoe: Yeah, I would say historically within this industry, eCOA, you know, we like to think we're all fancy and advanced as an industry, but in reality, eCOA, for a very long time, has just been basically paper on the electronic device, which brings benefits in regards to data quality, but really in in taking full advantage of the technology itself, it's not something we've been doing, and so notifications have historically been very static things.

You define them up front for all patients, they've seen the same message every time.

And you know, we've had good success in regards to compliance, but we truly believe that we could do better taking advantage of some of these new AI tools, and particularly our relationship with Click Therapeutics.

And so it was really exciting to see in the keynote, Matt talking about our initiative focused on notifications within the myMedidata platform this year, which is going to not only tailor the actual triggering of these notifications, as you said, so really learning actively from the patients when they engage with the technology when you see better compliance from them, so that you kind of support the patient in their natural environment, but also tailoring the message they're getting.

Maybe they need encouragement, maybe they need a bit of additional support, or maybe they just need a very data driven message of how compliant they are, or the benefits they're they're bringing to the study by being engaged.

So it's a very novel way of triggering those reminders for patients, and we're really excited to demonstrate the value that brings through better engagement, better adherence, better compliance to patients completing their tasks.

Aryana Hosseinkhani: Would your notifications be morning or evening?

Paul O'Donohoe: Neither lunchtime.

Aryana Hosseinkhani: Okay, so the big thing, the big purple thing, if you could see my shoes, they'd see their purple Dot, dot and the platform, I mean, all of our AI comes back to dot.

Tell me what we are doing with dot and interactiveness, with dot and Knowledge Hub data, connect all the cool things announced today. Where are we growing our AI capabilities across the platform

Sharlene Carnegie: Dot is meant to be all of the things ai. Dot is meant to be an agent.

Dot is meant to be a virtual companion. Dot is meant to be an orchestrator, right? Like we want to make sure that every opportunity we have to you know your point earlier on.

Redo. Use in manual, redundant tasks that you would typically do. Dot should be able to do that.

I think the fundamental shift that we're pushing here at Medidata is how a user interacts with our platform.

Instead of having a user navigate through multiple steps in order to do some specific task, we want Dot as a companion, as an orchestrator, to work alongside with our users on the platform, almost to augment themselves.

So imagine Dot you know identifying certain you know predictable actions that you may not necessarily have the foresight to take action on, but can recommend that for you, versus you having to remember, oh, I need to go take a look at this data discrepancy and manage, oh, I should probably look at this visualization, because it's probably going to derive some insight.

Instead of you having to go to that Dot should bring that to you.

And it's a completely different shift in how we think of the Medidata platform. It's something that is really innovative in our industry today, and it should change radically how people interact with the Medidata platform.

Paul O'Donohoe: It's almost a very obvious thing to say, but it's really interesting just hearing us talk about these different experiences so close to each other that, you know, we're basically, they're very unique use cases each one, but we're basically talking about the same thing.

It's about the burden. It's about not repeating task. It's about bringing things together, just for these different stakeholders and these different use cases.

Meghan Harrington: Visibility, transparency, so that people can actually make a decision knowing all of the kind of considerations, I think, is an important part, especially for the Study Experience.

Aryana Hosseinkhani: So most exciting question, I think the question you've heard probably a couple other times today, what do you dream to disrupt next?

I won't even pick on any of you who wants to speak up on what do you dream of disrupting next? I mean, we've already done so much with AI. What's next?

Meghan Harrington: Okay, all eyes on me i’ll go first.

I am really very focused my team, we're all very focused on trial design and planning, and I'm really excited about the disruption we think we can bring with the simulation, the collaboration across all facets of the work that takes place during design and planning.

Where I would love to go next. You know, I have a background in clinical trial financial management.

I've been hearing for the last 10 years the pain points in negotiating a clinical trial agreement.

I think that's a natural extension of the work we're doing right now with the digital protocol. What we're doing focused on study startup is, how can we really improve the negotiation, the exchange, the storing of those clinical trial agreements, because they're because they're a key part of the operationalization of a clinical trial.

So simulating and trial design, focusing, you know, kind of further out on what we can do with clinical trial agreements.

Aryana Hosseinkhani: To moving, so it is moving left.

Meghan Harrington: Yes

Paul O'Donohoe: I think I'm very excited about the idea of disrupting the inflexibility we've traditionally seen within eCOA and patient engagement with technology.

More broadly, I talked about the fact that, you know, historically, eCOA has just been electronic versions of paper, and I think within Medidata, we've always taken this approach of, let's build in the flexibility into the technology so that we can bring the industry with us.

They might not be ready. So originally, where designer was billed once decided to deploy across provision devices, BYOD web, the industry wasn't necessarily ready to fully embrace that.

And now we see the majority of eCOA studies, including some element of BYOD notifications.

The same we typically have very rigid, pre defined notifications. We're introducing this ability for flexibility and tailoring for the patients.

So I'm really excited to continue that effort of building the flexibility with control into our platform to really drive the industry to actually be meaningfully patient centric.

Sharlene Carnegie: I wanna call back on a statistic that was mentioned today that 90% over 90% of phase three trials fail.

And I also want to call back on something that you just said around the possibility with simulation today, we heard a lot about this concept of a virtual twin, or this concept of virtual plus reality.

I think if we do the right thing with the simulation capability, with virtualizing the trial, that statistic could completely be eradicated.

I don't think it should take us years to get there, right like clinical trials are extremely expensive.

They take a long time. It's a major investment, and we do need to radically increase the probability of their success.

So I think if we lean on this concept of the virtual twin, if we really focus on simulation, we can probably get close to 100% success rate with clinical trials.

So I think if we disrupt. Anything in the industry, for me, it would be seeing that vision come to life.

Aryana Hosseinkhani: So to close this up, if I was to summarize, we want to move less into execution.

Obviously, execution is very important, but move left. Be able to design the right studies up front, be able to scale that and have it be more configurable, more customized, personalized. Meet the patient where they are, and then ultimately, that will drive us away from this one out of 10 trials succeeding to four out of five trials succeeding.

So thank you guys so much. Sorry.

Meghan Harrington: I was gonna say, Sharlene said 100% so let's go. Let's go for that.

Aryana Hosseinkhani: We're gonna go instead of four to five, five out of five. So thank you guys so much for joining me today. I'll see you around here.

We'll be talking more about AI experiences and keeping the conversation going.

Meghan Harrington: Thank you very much.

Jeff Ventimiglia: Welcome back to from Dreamers to Disruptors. We are coming to you live from New York. We are here at Medidata next, and I'm actually really excited to be with the three of you.

We have Bob, Jia, and Bryant here. And the cool part about what we're about to do is that the four of us really do have this conversation on a daily basis, I would think so we're going to spend a little bit of time today talking about the introduction of AI into life sciences.

And I think I start with you Brian, because it's good place to start.

This is, it is not, you know, regular AI that's just sitting out there for for, you know, regular consumer use.

We are talking about life sciences. We are talking about patients.

CO can you talk to us a little bit about the scientific direction of the AI and maybe a little bit of how we think about trust within that?

Bryant Fields: Yeah, so I take a slightly different, well, different angle on that.

So reflecting on something I saw this weekend. I won't name the company, but it's a huge realty company, and they were really showcasing, in a very elongated way, their new AI platform.

So as a consumer going out to look at homes, you don't have to go out.

You can literally go inside the home, do in home review.

You can say, Oh, can you move the couch over here?

Like using NLP, change the color, change the color of the walls.

And I thought about, I said, you know that that's really cool. And thinking about that in the context of what we do on a regular basis, you know, it's not about getting the shade of the pink color to your preference.

There are real consequences. Some of those consequences, consequences can be, you know, a black box warning, and so really getting it right from a scientific perspective really matters.

Our clients come to us, you know, on almost a daily basis, and they're really honed in on the biomarkers.

They're really honed in on the the standardization of our data. They themselves are tackling data silos. So it's really important that, you know, the domain that we have, the domain experience, and that we can, you know, generate insights and models that really are fit for the purpose of what our clients are doing from a scientific perspective.

Jeff Ventimiglia: Yeah, before we started, Bob said, Don't make this so consequential.

So Bob, how are we building our AI that it is going to build trust and we are going to build confidence in the way that we've structured our AI strategy.

Robert Lyons: I mean, it's a great question. I didn't actually say it shouldn't be consequential. It's fundamentally consequential. It's the whole reason why we're here.

The I think the point I was making was that let's not start with danger. Let's start with opportunity, because we're at a moment of innovation and a moment of excitement in this field.

But there is no doubt that what makes AI possible is the humanity that we bring to it.

The fact is that we need to encode humans in the loop. Historically, I think, has meant guardrail or something that prevents us from taking the wrong step,

but we need to get to a place where we're encoding within our machines and within the intelligence that we build our software. I mean, I'm an engineer, so I think in terms of like doing that, as opposed to a process of humans actually being a checkpoint.

I think that ability to encode our judgment and morality into the software is something that we're getting closer and closer to be

Jeff Ventimiglia: Okay, it's understandable, but you know, John, maybe I'll come over to you for this. I'm on my mission to stump you.

I think Bob brought up a lot of really good points, but we still do hear, and I have a very hard time saying this word, but we still hear about hallucinations, right?

And we still have this concern. So again, you know, when we think about building, how are we thinking about repeatability and quality and accuracy of the data as we are? Again, bringing this to something that is very consequential in life sciences.

Jia Chen: Yeah, that's that's really near and dear to everything we do. How do we really is instill that trust in our product and to our users?

So I would say I was asked this question also early today, but to see the one way to look at this is that the impact we generate, right?

So, for instance, in collaboration with BMS, we were able to demonstrate a surrogate endpoint that really shortened the trial from 12 to 18 months, finding a strong 12 months, 18 months to 12 months, finding a very strong association between a complete response and the final the overall, traditional endpoint, like overall survival and the trial, got accelerated FDA approval.

So I think this is like, you know, with the regulatory approval, that's part of the trust?

Jeff Ventimiglia: Yeah. I think that the proof is the continuum of confidence builder, the more that we can do that.

Jia Chen: Yeah, but I do want to mention that, how do we put guardrails into the work we do?

An example I like to share is our collaboration around the virtual heart so the customer is asking us, if you take a hybrid approach between the physic- based the finite element of modeling to AI driven approach stimulants, how do you make sure the heart you generate is not going to be somebody from from aliens?

Jeff Ventimiglia: I knew you were gonna talk about aliens eventually

Jia Chen: Here we go. So we have a five layer framework, but basically looking at, how do we ground our AI in the laws of physics, biology, chemistry, bio mechanics, and also we check that with actual population in the real world and a clinical evidence.

So we have a multi-layered framework to ground ourselves.

Jeff Ventimiglia: I think you just named every class I took in college,

Robert Lyons: so it's fair to say you didn't stumble.

Jeff Ventimiglia: Yeah I didn’t stumble yet but we're not done. We still got a little bit of time here.

So I want to go back to the guardrails and pose a question I've been pondering myself, is we talk about human in the loop as a guardrail in a lot of ways, and the concept of building confidence, and I do believe that it's necessary as we jump off.

But is human in the loop going to be required forever? Let's get a little esoteric here. You know, will AI become good enough, and will the confidence be there where we no longer need human in the loop.

Robert Lyons: You want to start with me

Jeff Ventimiglia: Yes I was looking at Bob, and I forgot we were on film. So let's go with Bob. But I know Bryant has a response as well.

Robert Lyons: I think yours. I think I know your response. I mean, I've heard your response recently, and I think you should go second, because I think it's got a powerful, powerful.

Yeah. I mean, I already sort of previewed my position on this, that human in the loop isn't a scale limiter. It's it's the thing that brings legitimacy to what we're doing.

It's one of the things that brings legitimacy to what we're doing. You just gave along

Bryant Fields: Sure let’s do it.

Jeff Ventimiglia: before we get to Brian, though, I want to challenge. It's not a scale limiter like in the time and the effort that we put into human and move. Does that not like delay our scaling ability?

Robert Lyons: It depends on what you mean by scale. If you if the goal is to get to a good outcome, then no, it doesn't have to limit scale.

It certainly isn't. There's no doubt that ethics can slow things down. You can go faster and get to a bad solution. But that isn't our goal.

Obviously, it's it's getting to a place where, like the the judgment that exists in clinicians and researchers and patients themselves is part of the autonomous decision making process, and that in the same way that we're trying to endow our models and our intelligence with physics and biomechanics and all of the scientific principles, it also needs to be bounded with some of the philosophies of what we think fairness is and what goodness is.

Jeff Ventimiglia: All right, I understand that. And before Bryant tells us that we need human in the loop forever, though, but if we think about it, the reason that we are creating AI is because humans themselves are, you know.

Fallible, right? That they are the ones who are oftentimes making the mistakes as currently. So does human in the loop fix that problem?

Bryant Fields: Well, fallible. It's fallible for now, right? Today's failure is, is a foundation for tomorrow's success, right? No one ever gets it right the first time.

But my answer, the foundation of my answer, is that we're in the business of patients. Our mission is centered around patients.

And therefore we cannot escape the human. The human is the patient in this case.

So whether, like, if you just think about today, we have, you know, we go out, we get one of those, one out of 10 successful, you know, products that gets that does make it to market, then what do we do? We do surveillance.

Those humans. They're humans who are still informing, whether it's, you know, SAS code or or a model. Humans are constantly informing what we do.

So from my perspective, it's just a question of which direction the loop goes, right. Does it go back to a black box warning because, as we talked about, we failed somehow earlier and we missed something, or are we immediately looking to incrementally improve on what is now the new standard of care, you know, looking at sub populations, looking at, you know, looking at biomarkers to make sure we really, you know, have it, have it, right?

I'm looking at it from the patient.

Jeff Ventimiglia: Gotcha, I have spent a lot of time here at NEXT with our customers, and I often ask them, what does success look like when they're implementing AI, right?

And I do get a very common response, which I think is a good one. We want to get drugs to market quicker, but that also seems like an extremely lofty and hard goal to achieve.

So Jia, as we're thinking about the areas where we can implement and find success, at least to start with, where can we focus in within clinical research or in the life sciences space, where we can really start to achieve some of those, those builds and efficiencies?

Jia Chen: Yeah, I think today, early today, we talked about insight, simulation and workflow automation, if we want to kind of come down that level.

I think insights, I talked about how to glean from historical data to get to the endpoint faster, but also it's about patient safety, right?

So for instance, for T cell therapy, a trial, you only got 10 to 12 patients, so it's very difficult to understand under what condition the patient will go to severe adverse events, but with our historical data, we were actually able to find those predictors and take mitigative actions.

Jeff Ventimiglia: Got you so through the simulation, the insights to build things that we saw on the main stage. Today, we can start to condense those timelines from the source, right? Yes, cool.

Now, Bob, I want to, let's zoom back out again, right? Because we want to talk about the intelligence. We want to talk about the orchestration. How do we start to look at this more end to end, right?

To where we're not doing a use case by use case, but now we are truly using AI to run a trial.

Robert Lyons: Yeah, I mean, that gets back to, I think, the question of orchestration right, and how we bring technology to that process to automate it in a way that brings together all of the other issues that we've been talking about.

And the way that I like to think about that is this sort of wordplay of moving from intelligent orchestration that's all about humans building processes and defining an efficient way to conduct discovery, validate that and make sure that that the treatments are getting delivered as fast as they can, but still, ultimately a manual, Linear, reactive process to something that's more closer to what you saw Anthony talk about on the stage today, which was orchestrated intelligence, right?

It's the same words, but flipping them gives it a different dynamic, and I think that's where some of the new AI tools that we're bringing to our platform can really shine in accelerating the ability the system.

We're not just building workflows manually. The workflows are being reasoned about by the system at the time that they're executing and they're adapting and they're reacting with all of the science and ethics that we've encoded into them, and that is a pretty exciting place to be, and it will ultimately allow us to deliver, you know, things much faster.

Jeff Ventimiglia: Gotcha, that's awesome. I want to go back. We're on a podcast here called from Dreamers to Disruptors.

So I want us to be dreamers for a little while, like I want us to maybe think about what the what is five or 10 years from now look like that that maybe we're not talking about at the moment,

but when you guys, we all work in AI, we all, we all spend a lot of time talking about this, like, where, what do you think the wow moment is going to be when you dream about where all this is going, Bryant

Bryant Fields: Similar to where we were 5-10, years ago. I mean, in a different way. I mean, we heard this morning on stage from one of our speakers.

You know, the two things they want to see they want to see, you know, an increase, you know, in the probability of technical success absolutely, you know what we're doing, from an AI perspective, can absolutely impact that literally, almost run a trial before you ever enroll a patient.

There's so many things that we can do there. The other one, again, more patient centered, is around, again, expanding the use of synthetic data.

Like, really, I mean, I was plugging away at this for, you know, eight years, the first eight years here, you know, we're getting there.

We're getting there, I really think in the next, you know, five years, we, we really, as an industry, have to, you know, cross that threshold, yeah, and I think, and the difference is, is that I think back then, you know, when I first started as he saw Joshua, he's one of the people who interviewed me here.

We were asking me questions about big data, right? That was the big thing, big data, right?

And then it had the three V's. Then it became the five V's.

I think fundamentally, again, data is still going to be, you know, at the core of what we're, you know, the core in terms of the limitations as well as the opportunities that the opportunities we can realize. But, you know, ultimately, I think what we're seeing from, you know, seeing across the industry, with regard to, you know, folks coming out and really putting models out there to help really simulate what their true patient experience is, to the point where we can start reusing data or simulating the patients again before we even run a trial.

I think that's ultimately going to be big thing.

Jeff Ventimiglia: Yeah, yeah. So Jia, can you, I want you to dream big. You know, are we? Are we going to fundamentally change the way that clinical trials are conducted? What is it?

What does it look like to you in five to 10 years?

Jia Chen: Well, maybe even a longer horizon, I think maybe even go from sick care to healthcare.

I think by the time we get to the disease, it's probably been incubating for 10, 20, 30 years.

So the question is, can we go back to have really fundamental understanding of that biology and interaction with physiology, human body?

Jeff Ventimiglia: So so more, the more the intersection of Life Sciences and healthcare if you wanna move back into standard of care and be able, yeah, identify and treat patients earlier.

Jia Chen: Yeah, part of this is like a typical clinical trial, I think, still not very accessible to a majority of population.

So how do we really change that equation and and find a source of care and and health at much upstream than what we're doing now.

Jeff Ventimiglia: Okay. Bob, dreaming, dreaming big. What do you got for us?

Robert Lyons: I mean to me, it's just eliminating the the drudgery in the process, right? Like there's so much of what we do now. I'm a data guy. Spent the last several years at Medidata trying to build a better data platform, a way to bring data together, make it more accessible, more usable, like a world where you don't need to move data at all, and it's where it is, and you can use it in place.

That's something that I think we're going to see become a reality in five years.

It's happening now, being able to apply models, being able to bring together the best of what our partners are doing, and what we're doing dynamically and reactively, and to bring the quality of the signal up, but to also bring our ability to react to that signal in a moment where a patient needs it, I think, is where we're going to be.

Jeff Ventimiglia: I think, I think those are all really great points. I want to I'll close us out by moving out, move myself to disrupt her.

I think what's really cool is, when we have this conversation, Pascal is on stage. I've heard him talk about it a lot, is that you know 10 out of 10 planes that they designed fly.

And you know, how do we get to the point that you know 10 out of 10 drugs go on trial and are successful? And so I may not know the AI as well as you, but at dreaming big, like, how do we use the AI to, you know, synthesize the molecules, simulate the studies, build studies where we are planning for success, and really kind of making sure that we are not wasting time on, you know, failed molecules, and really focusing on the treatments that are.

Are, are going to, you know, bring the health care that those patients need.

So I want to see AI get, you know, really big and think about these things differently and not really be stuck in this milestone of phase one, phase two, phase three, but we're jumping right into it and know where we're going right off the bat.

So I think we are out of time now. So I want to thank my guest, Bob, thank you. This was an awesome chat,

and I want to remind our audience here that we will be back at this tomorrow. So please come join us on the podcast, or you know you can fly to New York and join us here next.

So thank you everybody.

Lisa Moneymaker: Hey everyone, and welcome to the finale of from Dreamers to Disruptors. Here live at NEXT 2026

I am joined by an amazing panel that combines Medidata and our Dassault Systèmes partners.

And so let's kick this off and as we bring it all home.

So Tom Doyle, Claire Biot, Patrick Johnson, I'm excited to have you guys here today.

I'm excited to be here with you. You're the experts in the area. Good times.

Thanks for having us all right. So as we're saying, we're bringing next to an end.

It's been two amazing days. Give me a little bit of a vibe. Takeaway.

So Tom, you've been on the floor, you've been talking to your engineers, you've been talking to the folks from our customers that have been coming together. So what's your one memorable moment?

Tom Doyle: I mean, it's been a great week.

I think the most memorable thing is just how much excitement there is for the possibility of the future.

So the AI, innovation, all of the work that really, collectively, our industry is doing, we're really starting to see how that's coming to life, how it's moving from concept phase into actual impact phase, like everyone is looking at like, what's next?

How can they have the biggest impact? And conversation after conversation, you hear almost that same thing. So that's really exciting.

Lisa Moneymaker: It's moved beyond, what if it's possible, or could it be possible to we showed it on screen.

Now, how are we going to get that in your hands? How are we going to do the change management? How do we do it faster? Faster?

That has been a regular occurrence, right? Yeah, directed at you.

Tom Doyle: No one has said, Hey, Tom, can you do that slower? That would be really great, you know.

Lisa Moneymaker: So Claire, you have been engaging with industry leaders. You've been, you know, out on the floor, I've seen you talking to pretty much everybody.

What's been the spirit that's really come across to you?

Claire Biot: I think it's obvious that we are the key catalyst for our customers to achieve precision medicine at scale with empowered patients.

And, you know, precision medicine starts with patients, and it's driven by science.

And we could feel this week that science is fueling our R&D solutions, that we're using sensors to define your digital biomarkers, using AI to de-risk trials.

But now precision medicine doesn't hit patients if you don't get at scale.

And we operate at scale, 8000 clinical trials as we speak.

And finally, we do this for patients. And you could hear so much about the patient inside board, and yeah, these patients inside board is actually very helpful in telling us these little things that could actually prevent patients from enrolling or sustaining the trial.

Lisa Moneymaker: That's right. And you know, that's one of the key drivers, is, how do we drive patient awareness of clinical trials as an option?

How do we drive the trust that's associated with that so they bring themselves to it? I think it's such a key factor in this I've been so glad to see you out on the floor.

So Patrick, you come to us from a science and a research perspective, and that has been a buzz in the other room at all of the booths.

What are your big takeaways? What are you hearing? What are you seeing? What do you want to make sure everybody leaves this event with?

Patrick Johnson: I think I have two takeaways, really and massively from the plenaries.

It was a reveal to me that we are the only company that really efficiently deliver, at the same time, an end-to-end business platform and a truly scientific platform for clinical research.

I mean, this connection from early stage down to manufacturing, this is a reality now what Tom demoed on stage with Anthony and yourself.

It was a reveal, because it was not only performance, efficiency on this, on the current processes, it was also catalyzing new processes, new ways of working, always, with biology, chemistry, physics, understanding of diseases with the patient at the center.

And if you can, if you, if you look at the customer reaction, I mean, it's more than adoption. It's excitement. It was intense.

Lisa Moneymaker: It was palpable.

I mean, you know, when you're on stage, and we had a lot of time on stage yesterday, there are some crowds that, you know, you give a lot to them, but they're not kind of giving it back.

Everything we gave out there, we were getting 10 fold back. You could feel the excitement in the room and, and I think you're right, it was a reaction to what they were seeing on the screen.

Like you said, that is real, that's coming to life. They could feel it. But you teed me up for something perfect, which is, you know, when we think about Dassault, when we think about Medidata, just.

Has been in the virtual twin space, in aerospace, in manufacturing, in automobiles, but now we are saying it is time.

It is here. We are bringing it to the human body. What does that mean for you, and how you guys are approaching research, what it means for us, and how we are tangentializing that

Patrick Johnson: it's very central.

And in fact, we've been doing and walking the talk without really saying it, and now we're ready.

A twin is really a 360 view of all data, knowledge and know how that you need, either to understand the mechanism of action of a drug, or to conduct a process of clinical trials, whatever complexity it is, or to manufacturing a bio produce the drug.

It's really the embedding, if you will, of all the knowledge now from the industry, which I guess we have accumulated with the more than 30,000 trials that you guys master.

So this is what we called virtual twin, and they rely today, generative, AI buzzword, on World models. Those are really the 360 understanding. And so we've been injecting that in the industrial processes.

And of course, we are injecting that in organs, in twins of the bodies, in cells, in tissues. And those clinical trials are going more and more to evolve towards simulation.

It's not so much a system of operation. It's a system of prediction to be able to optimize a protocol, for example.

And so those are exactly the science DNA that DS is injecting on the Medidata solution. And the twinification.

Lisa Moneymaker: twinification, yeah, well I think, you know, if you go back maybe two and a half three years ago, even when you're thinking about, is this going to be a reality and a possibility, like you said, of that, it feels palpably different.

Now the reality of, can we actually run trials in this way?

Can we actually run it through these, you know, virtual twin dupes?

I think what you're saying is it is here. It has arrived.

Patrick Johnson: twinify

Tom demoed, a trial simulation, a test population simulation to optimize.

This is a twin. It's not anymore only data analytics.

It is really a cockpit to steer your protocol design and to steer your trial execution.

Lisa Moneymaker: Do you see how he tied in the plane there? Yeah, that happened exactly. It was visible.

So, Claire, when we think about the end to end ecosystem, you know, we touched on yesterday, this idea of supply chain to manufacturing, how much does so plays a part in, you know, 80% of drugs that are FDA approved, 50% of medical devices that are out there.

You know, what's your perspective from industry on that?

Claire Biot: So I you know, today, clinical trials have the bottleneck to bring new drug to market, but the goal is to go twice faster.

Now, if that goal is met on the clinical development side, the bottleneck would be, how do you manufacture that drug at scale?

And so if we want the drug to be able to reach the patients, you really have to adapt the way you're going to produce your drugs.

And here we come into play by building the virtual twin of the recipe. To some extent, it's like cooking.

You need ingredients, equipment, and if I'm going to transfer that recipe to you, for example, a chocolate cake, you want to be able to feed the 1000s of people who have attended next.

And my recipe is for six people. So how do you scale this?

For this, you're going to use the virtual twin of your ingredients, your equipments.

You're going to see how it fits into a facility.

And we're moving from that time when you had one line for one drug in one site to modular facilities that you can rearrange like Lego bricks to adapt to precision medicines. Which means that you're going to have smaller batches for fewer patients and so wider variety of drugs that are produced in one site.

Lisa Moneymaker: I don't know that we're trusting him to cook. I haven't seen it come to life.

Tom Doyle: Just don’t give me any chef hats

Claire Biot: I share with him the virtual twist of my recipes, I think he's able to cook. No problem.

I'm gonna share with you not only knowledge, but know how

Tom Doyle: Quite a good cook. I will have you know

Lisa Moneymaker: Oh that’s fantastic. That’s good to know

Claire Biot: When are we invited?

Lisa Moneymaker: Yeah.

Tom Doyle: Well, that's a different story.

Lisa Moneymaker: You've never brought treats to the office. I'm just going to observe.

Tom Doyle: Okay, right now my limit is, but baby food, so

Lisa Moneymaker: Fantastic. So let's get to your you know, your day to day skill, which is making the interoperability come to life, the connection, the bridge between these topics we've been talking about.

What does that mean for you in terms of what you've seen today, but what your teams are working on?

Tom Doyle: Yeah, there's been a ton of work in how we can accelerate each of the domains that we are investing in, in early drug discovery, in drug development that comes from BIOVIA, of course, in clinical research and Medidata, and really elsewhere across the value chain, including in manufacturing.

What's really exciting is we're now starting to see a real opportunity to connect them in ways that were be very difficult to do before.

Some of that is the maturing of process. Some of that is maturing of technology.

A lot of it is the maturing of the virtual twins, though, and their ability to connect and interact not dissimilar. From now, the new ability of agents to enter, to intersect and interact, that start to connect bigger and bigger processes, that will be a lot of the focus as we look into 2026 and 2027

Lisa Moneymaker: Which of those do you think is the biggest accelerator?

Like, what's the one you're most excited about?

Tom Doyle: I think the one that's most that's going to be most impactful in the short run is the interaction of agents.

Is the interaction of virtual companions, the sort of connecting bigger processes. It's the one that people will see and touch and they'll feel and they'll see entirely new ways of interacting with technology that seem much more natural.

It's very likely that at the end of 2026 people will look back and think, I can't believe I ever used to work in that other way, like that.

That will come. But the biggest disruptor will be more what the what we understand from the underlying like virtual twins and data that's going to inform future research power, new medicines.

It's just those things will come later, and people won't necessarily know that they were derived, not necessarily from, like, what chemistry they came from, more in silico work and stuff like that.

Lisa Moneymaker: Yeah. So I'm going to, like, lead the room with this next one. We've got our research and our science perspective, our industry perspective, our technology perspective, you know, I'm going to go out on a limb and say, this gives us and our customers an unfair advantage if they're engaging with us in that way.

Tell me why you think that's so.

Patrick Johnson: Who starts it?

Well, the ability to explore, to do, what ifs the real purpose and the real utility of a twin is not to duplicate reality. It would be a digital twin. We do a difference between digital twin and virtual twin.

Lisa Moneymaker: Explain that, to the crowd.

Patrick Johnson: Digital twin is like a photo.

It's just a digitization in zero and ones of something. You digitize a document into a PDF.

A virtual twin is not that, because in a virtual twin you can explore alternatives, things that are not existing yet.

So you can explore in the virtual world, new ideas. You can explore alternatives. In the protocol design, you can explore modification and inclusion and exclusion criteria.

This is what we do, for example, in external control arm, they are synthetic based on real data, but you can explore and now, with the power of mod sim, you can do experiment treatment arm, which is a new way of using the virtual world and a prediction, but now with the pathology and the new drug candidate to evaluate the efficacy or the toxicity with simulation.

So this is what the we mean by virtual it's really exploration, what if alternatives to do, basically a system to conduct, to steer your study to whatever discovery, research process, development, tech transfer, clinical trials, manufacturing, options and so forth.

This is a cockpit to steer at the executive level and in all divisions in a company.

Claire Biot: I can bounce on that, because I think the second big value of the virtual twin, beyond simulation, is that it's going to radically change collaboration, you.

I mean, our customers suffer from silos, and it's silos horizontally.

How do you foster collaborations across different departments, but also vertically?

And so the question is, how can you from the boardroom to the decision maker, and how can foster decision making, and if something goes wrong at the time during development, let's say I'm screwing up with a batch.

How is it going to impact the clinical trial? Because I don't have any more supply. What are the alternatives?

So I really want to foster collaboration between CMC clinical manufacturing quality, and that's the power of virtual twin.

Lisa Moneymaker: It’s a hugely differentiating factor bringing all of those together. I mean, it's just a different way of thinking about how we can get from concept into patients hands. Yeah.

Tom your thoughts.

Tom Doyle: Yeah, I mean, first, I really like the way that Patrick described this. I would really like that photo that I can make myself a little thinner, maybe add a little bit more hair.

Patrick Johnson: Come to my lab. I also have

Lisa Moneymaker: Come to my lab he says.

Tom Doyle: But I think the part that we were, where I started, was what was most exciting about the vibe, is how much work and how much is already going on.

So I think of all of what we're doing, but also what all our industry is doing, and all of our customers are doing and together, that I think will really unlock some really new potential, and that should also be a big part of our focus.

It's not just the investments we're making, but the collaborations we're building to extend that out into a broader ecosystem, whether that's the agent to agent, whether that's bringing together more and more data to build a better 360 view, to build richer and richer virtual twins.

It's really a partnership, I think, opportunity for us as an industry to really unlock an entirely new era of innovation and research.

Lisa Moneymaker: Absolutely so on that idea of innovation and research and something new. So Patrick, it's cast your mind forward. It's 2027 we're back here. We're sitting at our table.

What's something that today? Is, you know, out of bounds tomorrow, in 2027 next year is just standard operating procedure.

Patrick Johnson: Protocol twins, protocol twins. You take a current definition of a protocol. It's a PDF today, and with AI and dot product we saw. It's almost almost there. It's next year.

You upload that in Medidata platform, and it generates a configured platform and system to operate and execute. So no need to go to it, no need to go the t use work.

It's from the document to data master and execution system ready to go.

So I think it's not that far, yeah, you say project, but it's next year.

It's already in the oven, and this is what I think will change the way people are looking at in terms of clinical trials.

Lisa Moneymaker: I was having dinner with one of our customers two nights ago, and you know, he was listing me his five things that he wants to make sure that we, you know, as a partnership, can make happen.

And his first thing was, you know, I think that, I think we can be building studies with AI. And I said, Well, youre gonna be very excited tomorrow.

And I'm also not sure why you don't know that we're doing it right this minute. But I said, it is a process thing now.

I said the tech is available, came out in November, like, let's get going. And I think you're right that by next year, this time, we're going to be wondering why we were even still building them.

You know, by hand.

Patrick Johnson: It's going to be this point, a next SOP, a next standard operating procedure for trials.

Lisa Moneymaker: Yeah, so that's a year away, but we know everybody's moving faster at this point, a year is too far.

Now, a year is too slow. Like you said, everybody wants faster. So Tom six months from now, what is the can't miss opportunity that you guys are working on?

But it's like this is coming. Be ready and start getting yourselves ready.

Tom Doyle: I think there's two of those. So one of the

Lisa Moneymaker: No I gave you one.

Tom Doyle: No, but I’m gonna answer with two.

Lisa Moneymaker: You could take two.

Tom Doyle: The first is, the canvas opportunity as the adoption, the really embracing of a new way of working.

The thing with innovation is it requires people to get excited, fired up, and turn it into something real, turn it into something in practice.

And that doesn't happen in a software development lab, that doesn't happen in 350 Hudson Street, that happens out at sites and out all across the world and at sponsors and CROs every day.

And so the faster we can get to like, really embracing new way of working, that we better be far along in the next six months.

I think the other that, from a technology point of view, should be there is everyone should be doing protocol optimization, and everyone should be doing AI study build for sure, there's no reason not to.

And everyone should be looking at new ways to explore their data, using using new like aI capabilities, whether from the Medidata platform or in collaboration with Medidata and others, to really like, unlock those secrets from data. That is very possible today.

Lisa Moneymaker: It is. It's possible it's here, and it's a must. Do you know, right before we came in here, we were talking to some folks from our Site Insights Board, and you know, one of them brought up that idea that at the sites, how often they get protocols that are not optimized,

that are the opposite of that, that is creating that additional burden on the sites that we talk about that they know is creating burden on the patients, that is slowing down the process, that is collecting too much data that isn't a part of what we actually want to analyze.

So I agree, if you are not working with a protocol optimization tool that is using a virtual twin that is ready to allow you to simulate what this could look like going forward, you are falling behind, and you're doing a disservice to your sites at that point.

Yeah. So we'll go to the group with the question that everybody's been getting this week, what do you dream to disrupt?

And we'll start with Claire, I don't think you've started us yet.

Claire Biot: So what do I disrupt next? I want to help humankind achieve sustainable access to quality care for all.

Why does it matter? Well, today, a third of the worldwide population doesn't have access to essential health care, and health care is not sustainable.

Health care expenses grow twice faster than GDP, and I think one way to sort this is to actually promote value based care.

So measure the outcome for given patient. Because you don't want to pay for treatment. You want to pay for a good outcome for the patient.

Why is it so hard to achieve? Because you need to bring the entire ecosystem together, the payers, the patients, the physicians, et cetera.

But that's why dreaming big, and I think that Medidata has definitely a role to play in the broader Dassault Systèmes, because it starts with being able to measure outcomes in a simple, trustable manner.

And who better than medidata can measure clinical outcomes in a scalable way?

Lisa Moneymaker: Amazing. I love it.

Patrick, what do you dream to disrupt. You may

Patrick Johnson: I have two

Lisa Moneymaker: You can also have two.

Tom Doyle: Ooo then I get three.

Lisa Moneymaker: Start thinking

Patrick Johnson: It’s a business model.

Well, I have to the first one is really, I'd love to do for the life science industries what we have been achieving successfully for the adjacent industries, which is moving from document-centric systems to something which is really, as I said, a cockpit to conduct and steer businesses with twins.

Today, clinical trials are mostly run on document, workflow and stuff like that.

And that was okay for the last generation of clinical trials.

But as we say, because of the the numbers of constraints, the fields of the burden, of the of the job, of the expertise, now we need to explore alternatives, optimize, and therefore you need simulation in the play.

You need big data understanding AI coupled in a very intimate way with a way to modelize and simulate the different options and prescribe the right one. Yeah. And and all the industry is begging for this, yeah.

And so time is now for the twin of sight, twin of protocols, twin of trials.

And so my My dream is to do what we've been doing for the automotive, aerospace industrial equipment, which is to help them move from a purely document centric approach to something which is a next generation steering cockpit.

Lisa Moneymaker: You know, it's interesting. There was a My favorite quote this week. We'll give you guys a little like insight into what goes on behind the closed doors,

but one of the quotes on stage was, no one goes to medical school to write documents.

That's it, and it's exactly right. How do we get them moved beyond it?

Patrick Johnson: And the second one, and still research is, of course, there's a lot of scientific challenges and stuff, but I do believe that as a complement to EHRs and EMRs that we still leverage, you know, the real representation of your health, not only when you're sick, but even when you're in good health, the lifelong engagement journey with twin as a proxy of yourself, as a representative of yourself, mobilizing the mod the same, or biology understanding, I think that's The next generation of twins for health.

And I think we even are going to go it's not embedded systems, like what we have in on watches and wearables, and sensor cloud is a very good offer for that, but we'll have embedded twins that we will wear and that will alert us if something comes and in a new prevention setting.

Lisa Moneymaker: I love it. All right.

Tom first flag us with how many, and then

Tom Doyle: I'm going to keep to one. I'll follow the rules.

Lisa Moneymaker: All right, fantastic.

Tom Doyle: I'm going to go from a different angle, though, as I think Claire and Patrick has really covered some awesome areas that really could be revolutionized.

I want to talk a little bit about the interaction with some of those, and where I think there is still more disruption to have, and that's in the way that we consume technology, and the way that it supports, helps, works alongside us, is there all around us.

This is changing at a very rapid pace, the humanizing of technology, the way is becoming more supportive a companion that works alongside you, that starts to work more autonomously. This is an area I think I really want to focus on of what does that mean, both in clinical research and beyond.

First, what does it mean for researchers? What does it mean for data managers, for everyone in the clinical ecosystem, but most importantly, what does it mean for patients, and how can we bring some of the new technology, whether it's behavioral science, whether it's new user interaction models, how can we do that that creates a much more frictionless or seamless experience for people?

There's still a gap between what we can build and how we like to use it, and I think that deserves some real closing.

Lisa Moneymaker: I love that. That's fantastic. You guys have been amazing. We're coming to our end now.

I want to thank Claire and Patrick and Tom for joining us as we say goodbye to our audience.

From next, our final episode of from Dreamers to Disruptors. At NEXT, I want to remind you, and this is the first time I have ever said this in my life.

I'm very excited for this. Please subscribe on Spotify, Apple, YouTube, your podcast, listening device of choice, be a part of it, because we are speaking to industry giants.

We speak to, as we said, disruptors, people are thinking about the next big thing.

Please join us at next, the next year, or at any of our global events, but we've loved having you here. We've loved you being here with the crowd. The live event is really, really neat, but we want to thank folks, and we will see you later.

But thank you and goodbye from New York. Thank you.