Snowflake Summit 2026 | Dayne Turbitt, Snowflake & Emmanuel Frenehard, Sanofi

Clips
More from Snowflake Summit 2026

Dayne Turbitt

Senior Vice President, EMEA

Snowflake

Emmanuel Frenehard

Chief Digital Officer

Sanofi

play_circle_outline From Experimentation to Transformation: Building an Agentic Enterprise on Snowflake with Trusted, AI-Ready Unified Data

play_circle_outline Sanofi’s AI-First Strategy: Simulated Clinical Trials and AI-Driven Drug Discovery to Speed Development and Outcomes

play_circle_outline Cultural shift, upskilling, and large-scale employee engagement at Sanofi

play_circle_outline Rapid innovation in LLMs and competitive landscape among AI and SaaS vendors

Info
Transcript

Dayne Turbitt, Snowflake & Emmanuel Frenehard, Sanofi

Dayne Turbitt

Senior Vice President, EMEA Snowflake

Emmanuel Frenehard

Chief Digital Officer Sanofi

In this interview from Snowflake Summit 2026, Dayne Turbitt, senior vice president of EMEA at Snowflake, joins Emmanuel Frenehard, chief digital officer of Sanofi, to talk with theCUBE's Dave Vellante and Rebecca Knight about how enterprises are moving from AI experimentation to governed, production-ready agentic intelligence. Frenehard frames Sanofi's AI-first strategy around a single imperative: compressing a 10-to-12-year drug development cycle, where 90% of drugs still fail at phase three, down to five or six years through simulation and data-driven preci... Read more

explore Keep Exploring

How can organizations avoid creating fragmented, department-level "islands of intelligence" when adopting AI, and how can consolidating data into a governed, AI-ready platform (e.g., Snowflake) enable enterprise-wide AI workflows that truly transform work? add

How will AI-driven drug development change the long (10–12 year) drug development process and patient outcomes—for example, can it detect failures earlier and halt them, prioritize certain symptoms or targets, and use simulations (including virtual models and trials) to shorten and improve clinical development? add

How would you describe Sanofi's culture in relation to AI and digital transformation? add

How do you see the emergence of large language models and AI affecting enterprise automation and the competitive landscape (e.g., RPA vendors, Snowflake, Salesforce), and what is your company's strategy in response? add

bolt Powered by CUBE AI

Dayne Turbitt, Snowflake & Emmanuel Frenehard, Sanofi

search

Rebecca Knight

>> Good afternoon everyone and welcome back to theCUBE's live coverage of the Snowflake Summit here in San Francisco. I'm your host, Rebecca Knight alongside Dave Vellante, the co-founder of theCUBE. I would like to welcome our next guest to the show, Dayne Turbitt, Senior Vice President of EMEA at Snowflake. Welcome.

Dayne Turbitt

>> Thank you. It's great to be here.

Rebecca Knight

>> And Emmanuel Frenehard, Chief Digital Officer at Sanofi. Welcome.

Emmanuel Frenehard

>> Thank you.

Rebecca Knight

>> So I want to start with you, Emmanuel, because Sanofi has been very vocal about its AI first strategy. Beyond the slogan, what does that actually mean in practice in terms of how you operate as an organization?

Emmanuel Frenehard

>> Sure. I thought maybe a bit of context because important. In our industry, a drug discovery is a cycle of 10 to 12 years. If you're a patient in waiting, it's a lifetime. And so when we look at that, we say, what could we do to reduce that time and make medicine more accessible? So when you ask the question, what does it mean? Every step of the process, whatever we can deploy AI to gain time for the benefit of patients, we'll do so. From the way we discover drugs, to the way we take them to clinical trials, to the way we manufacture them, to the way we commercialize them. All that cycle, we want to improve it and we want to make it time efficient.

Rebecca Knight

>> So it's all about speed. And from when you hear Emmanuel describe that, what that means at Sanofi, how does that match the patterns of what you're seeing across EMEA and other countries and organizations that you're working with?

Dayne Turbitt

>> I think very similar. What I love about working with Sanofi is the personal mission around patient care. So it really identifies the gaps in a process that AI can automate. But if I look across industry, everyone is looking for that. And there's a big shift today from AI experimentation to how does it truly transform the nature of work? And that's what we're seeing everywhere. And what Emmanuel is driving is truly on the cutting edge of that transformation. How do I reimagine my business processes leveraging AI to make them go faster and be more efficient?

Rebecca Knight

>> And save lives.

Dayne Turbitt

>> And in this case, save lives for sure.

Rebecca Knight

>> Yes.

Dave Vellante

>> I want to ask you about how you're seeing this play out in the real world, because Emmanuel, on stage last night, you talked about... You started with fixing fragmented data and you said... That's the trivial part, not trivial, but that's necessary, table stakes if you will. And then you talked about the early value came from working with the forward deployed Snowflake team and then moving from untapped data to department level AI workflows. So there's a big debate going on, not say debate, but a caution that we're going to automate islands of intelligence. And so how do you avoid that and how do you see in the real world that playing out? Because what we're seeing is you have this top down mandate, Michael Dell, thou shalt be AI native. And then all of a sudden I said, okay, well, let's start playing with OpenClaw and NemoClaws and everything else. So now you have this personal productivity boom. How do you make sure as a data leader, an executive, that you don't just create all these other stovepipes?

Emmanuel Frenehard

>> I love the debate. I think that's a great debate. First principle, all our data shall be on Snowflake and shall be what we call it AI ready data, which means it's governed, which means of quality we can trust. So it's one environment. If you think about it, we centralized it. And then the second view that we have is companies are all different. Even in our industry. AstraZeneca, Novartis and Sanofi have similar goals, but we don't run the same way. So we want to be able to run workflows that are uniquely designed for our company straight on the data. No need to have intermediary systems. So the data is the gravitas. The workflow is what runs on top of the data. And so with Snowflake and the ability that... Snowflake is an operating system, it's not just a data platform. It truly is an operating system. So we run workflows straight... With Elementum and Snowflake, we run intelligence. So we're not fragmenting the intelligence because the data is unified .

Dayne Turbitt

>> So let me add to that. I agree with what Emmanuel has said. What's interesting about this time is we've gone through this period of tech rationalization where you've moved data from legacy systems to modern systems. What's different about now is there's a new reason to do that. And the reason to do that is in the way you want to change work and workflows. However, there's still fundamental piece, you have to get your data in a governed way. It has to be under management, has to be under governance and has to be trusted because if it's not trusted, you don't trust the outcome and the results. So what I'm seeing in the market is a renewed surge of connectivity through open systems or ingestion of data into Snowflake as an operating system. Then I'm seeing clients that have been doing this. You talked about a trivial exercise, it's certainly not trivial. However, it's been done at Sanofi. It's been done at other companies like AstraZeneca. Now the question is, how do I exploit that data in a way that I can reimagine my process? Enter in the agentic enterprise where you now start to treat agents as coworkers, you can start to reimagine what the workflow is. So if your data is governed and trusted, you can then unleash these agents to go and automate workflow tasks. And once you do that, you shorten the amount of time work gets done. So that's what I'm seeing. And different organizations are at different levels of maturity. Some still have highly fragmented data and are in the process of getting it consolidated. Others have a consolidated data system like Snowflake and are now in a position to exploit that with all the AI that's out there.

Dave Vellante

>> And to the extent that you can demonstrate that to the business, you're going to have a flywheel effect on Snowflake. I mean, you're not going to stop shadow AI in companies, but to the extent that you can show that the activity on Snowflake is going to drive business value faster, that's like gravity will create more usage.

Emmanuel Frenehard

>> It's a magnet. It's a real magnet for innovation.

Dayne Turbitt

>> Well, and what's great is... I mean, you've seen the shift of AI. Every executive is interested in doing AI. Then you have a chief digital officer that has to do the AI and it's two different statements. But the reason that you're doing it is that there are true business outcomes that are now possible. You're seeing the business outcomes that you're achieving. We're seeing many examples where they're putting AI to work on the crown jewels of an organization, largely the structured data, and actually getting real business benefit out of it. That starts to then fuel more investment and more re-imagining what's possible.

Rebecca Knight

>> Emmanuel, I want to go back to what you said at the very beginning of this conversation, which is that drug development takes 10, 12 years. For a patient, that's a lifetime. Can you talk a little bit about what AI-driven drug development looks like in this new era in terms of is it understanding the failures quicker and therefore abandoning them? Is it seeing working on some symptoms, not others? How is this actually changing things for those patients? It's

Emmanuel Frenehard

>> A great question. A cycle of drug development is very lengthy because you have the hypothesis phase of research and then you've got a lot of experimentation. Once you pass the experimentation which cycles, you end up with human trials and those can be many years. What if we could simulate all of that before we even run it? The sad story of our industry is that 90% of drugs still fail at phase three of a clinical trial. That means at the end, you're almost on the finish line. Patients are waiting and you fail. We believe, I believe, that we'll be able to see and we're starting to simulate even clinical trials before they happen. So we still have to do them, but we do them with more precision. We do them with a better appreciation of the kind of patients we need in those trials. So my hypothesis that you'll simulate molecule making, you'll design them, you'll simulate a molecule's impact on the targets. Could be a cell, could be a gene, all of that will be... You'll simulate it in mice, mice models, virtual, twins, organs. You'll get to humans. Human biology remains a major mystery for all of us, but we're going to get to a place where we'll simulate. And my intention is to have the time it takes. So from 10 to 12 to five to six.

Rebecca Knight

>> Wow. Wow.

Dave Vellante

>> When we went from on prem software to SaaS, the operating model changed, the technology model went to cloud, the business model, the pricing changed to consumption. Companies like Snowflake were obviously born and affected, but it was really the IT department that was impacted. Now, as we go from SaaS to AI software, everything's changing again. Forget the technology model for a minute. We're all moving in that direction. I'm interested in how... And we're just trying to figure out the pricing model. We're talking about outcome based pricing, but how do you see the operating model at your company change? Because it's not just going to affect the IT department, it's going to affect every organization, every department across the board. How are you thinking about that?

Emmanuel Frenehard

>> That's probably the hardest. That's the hardest thing we have to solve right now. I think we've entered an era where we can custom design processes for any company, any function in the company. And you see, agents are here to do things. But to do that, I think human nature is to optimize what we know. We're very good at shaving time, reducing friction, eliminating steps. If I give you a blank sheet of paper and say, "Design to me a procurement function of the future," you can have a blank. But if I tell you, design it with the hypothesis that you have agents, that hybrid workforce can help you. We are entering, I think, a true era of reinvention of processes where it's not just the IT department, it's designed together. The IT department will code the agents and will maybe own the agents, but the logic has to come from within the company. And that's probably where we need still a wave of, I guess, acculturation and upskilling within our, at least even our organization.

Dave Vellante

>> Any thoughts on that, Dayne-

Dayne Turbitt

>> Yeah, I think what you'll find, there's a concept of expert generalists. So you used to have the experts that were deep in database or they were deep in some coding level. With the invention of tools like CoCo or Cortex Code, you have an agent that can code for you, okay? I'm sure you've heard of vibe coding, but let's talk about that. So now you have people that are expert generalists that are able to interact more intelligently with the IT department because there isn't this technology barrier anymore. And to Emmanuel's point, it's the reimagining of the business process that's important and you need both the technological knowledge and the business knowledge and I'm finding more joint teams being formed. Much closer integration between a business user owner and an IT professional so that they can go and reinvent what we want the agent to do. We see that internally where we use a product called Raven and we're having a challenge now on how do we reimagine forecasting? How do we reimagine how we deal with our clients, not on the basis of our current paradigm, on the basis of, well, what if we had agents that could do research, that could draft emails, that could analyze Sanofi's current use case and then suggest the next best use case based on the other 60 pharma companies that were there. All of that work makes my Salesforce way more productive. That took a lot of collaboration between the sales organization and the IT department just to document what they do on a daily basis and how to automate it.

Dave Vellante

>> So if I may, pre-Raven was probably good at what happened, maybe a little bit why it happened, but not great and really deeply why it happened, what's going to likely happen next, what's the next best thing that I should do? And that's what Raven presumably is designed to do.

Dayne Turbitt

>> And that's where we're going from a sales perspective, is if you think about the amount of time as a sales professional you spent researching a particular problem or a particular client, it's all productivity. Now I have a technology or an agent that can do that research for me and present me a hypothesis much like you did today or yesterday on stage, but that then suggests next best action based on what we've seen from our 12,000 clients, which I don't have access to, but the system does. Then all I have to do is go and take that to the client and work on that client's particular business problem, but I arrive more educated, more informed, and it's a far more intelligent interaction with our clients than when I just show up with some random product that I'm trying to shift. So hopefully it's a better experience for you.

Rebecca Knight

>> I'm thinking about the cultural shift inside of organizations based on what you were saying in terms of the technologists working more closely with the commercial team and at a place like Sanofi, it's also the clinicians and the scientists. What has been the hardest parts about getting teams to work in new and different ways and what maybe has surprised you, Emmanuel?

Emmanuel Frenehard

>> We had a few strong messages and not everybody agrees with them, but AI is not for automation was one of our messages, to do things that don't digitize the past and try to AI fight the past by saying, "Oh, with an agent, I'll do it faster." No, truly rethink what the future could be. That was a strong message, which in the beginning I think was not fully understood. But when you start opening the land of opportunities that this provides you, it changed the dynamic. And so I would say Sanofi is a place where we have tremendous excitement for it. Every year we run, for a month, we run what we call digital months. We do workshops, hands-on workshops. Snowflake came and presented at it. We have 75,000 employees at Sanofi, think about people in manufacturing and 44,000 employees attended those workshops. I've never seen that. So when you ask, how is the culture? The culture is of innovation, of willingness, of daring, of risk-taking, all of these things and willingness to learn. And so all the ingredients are here.

Dave Vellante

>> Yeah, I call that paving the cow path. Don't just an existing process. I mean, it pains me because we love our friends at UiPath, but RPA was really what it did. I mean, I presume you're familiar with Cortex Sense. That starts to get into the process level capabilities. It's early days, but as I've been saying, three or four years ago, Snowflake wasn't even in that conversation. And now you're right in the mix of that interesting competitive dynamic. You got the LLMs over here, you've got the SaaS companies. I saw a sign today, Salesforce. So they've got great affinity with Salesforce, but they would love to be that system of intelligent, of course, Snowflake as well. So you've got a really interesting cage match, I think we call that going on here.

Dayne Turbitt

>> It's definitely an interesting time in the market. The LLMs have had amazing success. You saw the Anthropic success story on stage and whether it's OpenAI and whoever's next. I loved in the keynote with Sridhar that they spoke about, that was five years ago, okay? It felt like the dark ages. Things are moving so quickly and we'll see how things play out. I think going back to Sridhar's point, we're very focused on delivering successful outcomes to our clients and doing what we do best, which is providing a governed trusted data platform. And once you have that in place, AI just accelerates that.

Dave Vellante

>> And keep innovating on product. I said this before, Sridhar said product market fit is a life force. So keep innovating. And that's true for any business that develops-

Emmanuel Frenehard

>> Any industry, any company.

Dave Vellante

>> Right. And so that was an homage, by the way, to Frank Slootman.

Dayne Turbitt

>> If you saw Christian's keynote today called it an avalanche of innovation.

Dave Vellante

>> Right? Yes.

Dayne Turbitt

>> And that's certainly how the last 12 months have felt. So continue, which is great. I think it's super exciting.

Rebecca Knight

>> Exciting times. Dayne and Emmanuel, both, thank you both so much for coming on the show-

Dayne Turbitt

>> Thank you for having us.

Rebecca Knight

>> A really interesting conversation.

Emmanuel Frenehard

>> Thanks guys.

Dayne Turbitt

>> Thank you.

Rebecca Knight

>> I'm Rebecca Knight. For Dave Vellante, stay tuned for more of theCUBE's live coverage of the Snowflakes on it after a quick break. You're watching theCUBE, the leader in enterprise tech news and analysis.