DRUID Talks Ep #10 Tech-Driven Pharma: Pioneering Conversational AI for a Better Life
Watch this episode to discover real-life examples of how Servier Pharma's innovative use of Conversational AI sets it apart, reshaping how pharma operates.
Episode #10 of the DRUID Talks Podcast features Laurentiu Bogdan, Digital & IS Director at Servier Pharma and Subject Matter Expert Kieran Gilmurray. See the full episode and transcript below.
Kieran Gilmurray: 96% of Servier Pharma employees use the Conversational AI assistant every single day. That means over 10,000 conversations are happening each and every single day inside of the company.
That means, Gelu, the chatbot, as it's named, has a voice in Seriver Pharma company. It's a 4.2 billion-a-year company with 22,000 employees. And this has been made possible by one of the most innovative or cost-effective IT teams that I've come across in ten years.
Welcome, Laurentiu! How are you today, sir?
Laurentiu Bogdan: Welcome! Fine. I'm very well. I'm glad to be here. Thank you for the invitation. And I hope we'll have a nice conversation. And I'm pretty sure that will be nice because we have a lot of good topics these days, AI being one of them.
Kieran Gilmurray: Well, let's share those topics. But, everyone, let's jump straight into them. Let's start at the very beginning.
So, for those people who don't know Sarvier Pharma and what makes it a big name in the farm industry, could you tell us a little bit more about the company?
Laurentiu Bogdan: Sure. So, yeah, it's not really a big group. We are big, but not the biggest one. Servier is an independent French pharmaceutical group. And it was founded in 1954. And what is important to highlight is the fact that the company has a passion for innovation. And we like that; it's part of our DNA, and we want to improve the patient's life, of course. We are present in, I think, 154 countries. And another thing which is important to be highlighted is the fact that our governance model is what makes us different and also what makes us strong as being governed by a foundation. We don't have another way. We are not paying, you know, to different shareholders and so on.
So, because of that, we can, you know, of course, in a sustainable way, sustain our innovation and improve the patients' lives. Yeah, that's basically who we are. We are proud of this. I am proud to be part of this company because of that because we are working for patients, for real, you know? We are not paying; we are not working for our shareholders.
Kieran Gilmurray: Oh, I wonder if that's changing how the company operates. But you're being very humble. And that's how I got to know you. By the way, as part of our interviews when we were preparing. But there's something about Servier Pharma that many people don't know, and that's the IT team. And this is the bit that I want to take in. Laurentiu, you said when we were preparing for the interview, and this was quite different for me; the role of IT is to seamlessly build apps for business, the people inside it, that they don't know that they need, but they will never give up.
Is that what inspired you to implement a Conversational AI megabot? And what's it doing, and what needs to say/respond to all these wonderful people inside of the company?
Laurentiu Bogdan: I think the role of IT, in general, should be around the people. Because you are not there to serve technology. You are there to serve people. And that's the power of the technology. You can make their life much easier because this is about technology, the efficiency that technology brings; it doesn't bring for the sake of technology but for people. So, if we are talking about the people, then each tool that you want to implement, you have to implement, having in mind first the people and then the technology.
Megabot. Why megabot? Because if you are thinking about people, you have to have something which can answer all the questions that you may have from one place, one mega-AI, whatever you call it, megabot. Imagine if you have one person in the company who can answer all the questions that you may have, no matter what the topic is. I think that's the real benefit for people, not for technology.
Kieran Gilmurray: I love that! That was the bit that attracted my ear when we were talking because normally, companies just implement an HR chatbot, they implement a finance chatbot, they implement an operations chatbot, which then recreates the silos of an organization in every sense: in the people sense, the structural sense, the technology sense. So, I love this concept and this idea that you have the most knowledgeable person/thing inside of the organization that knows everything.
That must not only please people because then they can’t get an answer without moving from department to department, but there must be something in that Laurentiu that says, „If I can incorporate all of the institutional knowledge inside of the organization, that it doesn't matter when people come or go or leave or move, it's available at people's fingertips.”
Is that how it operates in the real world?
Laurentiu Bogdan: Yeah, I think yes. And also, I'm thinking now we have heard the "collaboration" word many times. We have to collaborate better between departments and between different actors in a company. Now we have the chance to do that because you have the technology which can do that. It's, in a way, stupid to maintain the same silos when you can do it simpler in the only way so you can achieve what you are talking about for years very simple.
That's why I'm saying that it's a very good chance to achieve what everybody has wanted for years, you know, real collaboration.
Kieran Gilmurray: But that was the interesting bit, Laurentiu because people have talked about this for years. And for a moment, during COVID, everybody in the organization worked together. But I get the impression they're now going back to their old ways of working. But you've deliberately rethought how IT should actually function. And not only that, if I ask you the next question here, at the same time, you've rethought all IT functions, but this is at a time when budgets are not rising, budgets are actually shrinking, incomes are falling down, and costs are rising up.
So how did you meet the challenge of actually implementing Conversational AI and all of the technology changes that you've been able to do on the same budget that you had more than ten years ago?
Laurentiu Bogdan: I think, yeah...You know, I don't know how to make it to be clear, but imagine that you have, let's say, the same amount of money for many years, but you are not doing the same thing every year. So, if you have to, if you are planning your budget to have every year some money for innovation, you will have it, it’s easy. You are innovating now, you finished that innovation, and next year you use the same amount of money for the next innovation that appears.
So, you don't have to invest more and more just to have an innovation. I don't know how to explain it very well, but I think it's not about the money. It's about the will to have, to find a way to use what is new and what is useful for your employees, in the case, you have projects for employees. What do I mean? You can keep the budgets fixed. Just you have to plan better what you are doing with them.
Kieran Gilmurray: I love the way you say that so simply. And that is... this was the beautiful bit about preparing for the interview. It’s the most natural thing for you to say, but I see year after year IT departments demanding more and more and more budget to innovate. And they deliberately don't innovate if they're not given cash because then they get to say to the leadership team, "But you haven't given me money; therefore, I can't do these things."
But you've been able to do it without seeing rises in your innovation budget or your tech budget. That's the bit that's extraordinary. I love the way it's so simple for you!
Laurentiu Bogdan: I think it should be simple. You have a salary. It's the same salary, doesn't mean that you are not innovating in your house because you don't have more money to do it. It's the same thing, in fact.
Kieran Gilmurray: I loved that. We need you in so many more IT departments! CFOs must love you!
Laurentiu Bogdan: Yeah, probably.
Kieran Gilmurray: Here, let me ask you a question then. How do you convince people in Servier Pharma to use Conversational AI assistants?
Laurentiu Bogdan: There's perseverance and some tricks to do that. So, from my experience with employees and with internal clients, it made me think that sometimes, most of the time, you should first implement it, let it in the field, and they will start to play. Then they will understand the benefits. Of course, I say it in a simple way, you cannot throw any kind of project in the organization just to see if everybody wants it. It is very important to know the business. If you understand each part of the business, you will have an idea of what can work and what cannot work.
So that's why I said many times that an IT department is a good department if it is 100% in the business involved. Because in that way, they can understand the business well. I was afraid to say” very good” because you cannot understand „very good” everything, but at least you will understand well what the needs are, and then, I am pretty sure that you will find the right tool that you can put in before they want it; because that's another philosophy that we have here inside.
Our goal is to bring to the people tools that they don't know they need before they ask it. Because if they are coming to you and ask for a tool, you are too late. You are not a consultant, a technology consultant anymore. You are just a supporting department. And they hate that tag that the IT department says, „IT is a support department”. It’s not. If you are looking at the world now, everything has something digital behind, everything!
Kieran Gilmurray: See, there are a couple of nuggets in there that I don't want people to miss, and this is the bit that I just adore. Laurentiu, having talked to you, because you saying this is the most natural thing in the world. But I come across IT departments every single day that are classic IT departments; they're order takers, and then they wonder why the business treats them like order takers.
And I’ve seen a survey recently that says, "Is IT an enabler or not?" And over 67% of people said „IT is not a partner”. But you have put people in your team who spend deliberate amounts of time actually to understand the business. You put technology in place before the business even knows that it needs it. That serves a purpose that they didn't even think they had in a way that delights them, and then they wouldn't take the technology back out.
And you're doing all this using the same budget that you had ten years ago. Those things are very unusual.
Laurentiu Bogdan: Probably. What I'm trying to say is that it's in our hands (I am talking about IT) because I heard this annoying voice that „IT is not an enabler”, „IT doesn't do that”. It is in your hand to do it, if you want, of course. You have to be part of the business. You know, the business, it sounds like a separation, but that's the, in general, this is the reality. The business will not take you on the table and say, „Now you are part of the business.” No, it's not like that.
You have to be there. And to be there, you have to learn the language; you have to understand their needs, to really understand their needs, not the technology needs. In the first place, you have to understand the real business needs, and then the technology will come.
But it’s, you know, it is in our hands to do it. Of course, it's easier to stay in the server room because you don't have to deal with different personalities, with different... you don't have to communicate because the servers are very simple to communicate with. But if you want that, don’t complain.
Kieran Gilmurray: Oh, I see so many teams complaining. They need to copy what you're doing here. Well, look, you've put in the technology; you've convinced people to use it without them realizing they even needed this.
But what are the benefits your business teams have noticed since putting in Conversational AI to your customer-facing side of things and your internal organization? And how do you measure the benefits of your conversational assistant?
Laurentiu Bogdan: It is very simple. If we talk about only one single thing, you need a holiday. So, we have, of course, we have our digital platform where you can request the holiday, but to request a holiday through that platform, you have to open your laptop, go to the website, enter your password, fill in the data, and so on; it will take time.
What Conversational AI brings, it brings technology closer to the user because all you’ll have to do to ask for a holiday is just to open the Teams, because we have it on the phone, and say it in a natural language „I want a holiday. When? Tomorrow." And the Conversation AI will do it for you. And that's it. It is, you know, we saved time for the employee and for the approver because the approver will sign the holiday through Conversational AI in the same way: „You have a holiday request. Do you sign it? Yes.” And that's it. It's natural; it’s like using a phone.
Of course, if you are in the figures and you want the data to see how much money you are saving, it's not all about money. And, of course, we can prove that time means money. You know, we can talk about that, but it's I think it's more important to keep in the middle the people, the employees. And if you have few, if you give them the technology, closer, now, if you make the technology invisible for them, it is a big, big win because this is what happens. They will not see the technology as something that they have to learn, to remember what is the button or... You have to make the technology work for them!
And Conversational AI it's a very good example because it works on their language, on their way of working with speaking.
Kieran Gilmurray: You know what? There are a couple of lovely things about what you're saying, Laurentiu... I was talking to Nevenca Doca recently, she was on one of the podcast talks, and you and her, by the way, mentioned people at the core of everything that you're doing.
Because, and you mentioned as well, there's subtlety in what you're saying that I do want the audience to miss. You know, the words that I heard there were talking in the language of people, in allowing conversational AI to do that in the most natural way possible. Because, again, I see too many companies putting in this application or that application and this application to do this small thing. And that results in everyone having multiple passwords and having to remember how all these applications work, which always results in service desk calls or frustration, bad data, or something else.
But your megabot now in bearable language, in words that people actually use, can help in finance, HR, operations, and everywhere else. And as you said yourself, it's now become invisible to your people. That surely is the right way to put in technology.
Laurentiu Bogdan: Yeah. And I think we are in the right moment in time to build that because nowadays, with Conversational AI and the new language models, we can really bring the technology closer to them, and we can really make the technology invisible to them. It's, I think, it’s a great period for all the ITs in the world because finally, we can deliver something that can really change life.
Kieran Gilmurray: Oh, I wonder if IT departments want to become invisible. That's the interesting bit.
But what would your advice be...
Laurentiu Bogdan: Just a moment, just a moment. We want to be invisible forever because we don't want to be part of the business. That's why we're talking about that collaboration, staffing, enablers, and so on. But it's a, come on, you cannot say that you don't want to be invisible, but you are already invisible. You want to be part of the business, but you aren’t part of the business. Come on! You have to decide!
Kieran Gilmurray: One foot in, one foot out, I think. Well, Laurentiu…
What advice would you give to organizations thinking of investing in Conversational AI?
Laurentiu Bogdan: They have to start playing with the Conversation AI as soon as possible because it's really a game changer for clients, in general. I think you just have to start playing with it, and then you will see the benefits. Probably it’s not easy to do because you have to find the right processes that you can use. But once you start to build Conversational AI, you will find very fast the best use cases that can improve the employees' or clients' life.
Kieran Gilmurray: Well, if you were to mention use cases, what examples have worked really, really well for Servier?
Laurentiu Bogdan: Everything that is easy to automate, for example, holiday requests, expense notes, I don't know, I think I can put it in a general way, everything that is a request, it works very well with Conversational AI. And now, because we have GPT and language models, new AI language models, you can put the entire Q&A in your Conversational AI.
That’s another thing that I saw in time; in a company, it's very difficult to find the right information or the right people to help you with something. And if you are able to build a bot that can give you exactly what you need, you will save a lot of time. Otherwise, you are moving from one person to another, from one department to another department, to find what you are trying to find. That's why, now, with this combination of GPT and Conversational AI, I think finally we can solve that.
Kieran Gilmurray: Have you used ChatGPT integrations with your conversational assistant? And if you are, what are you seeing, and what are your expectations going forward?
Laurentiu Bogdan: We are playing with it; we are working these days, in fact, with the integration with ChatGPT, the Microsoft Azure OpenAI, I don’t know what is the name of the service, and we are trying to achieve exactly what I am saying, we are trying to have there the entire knowledge about internal procedures, and then we'll build on that. We also want to include competencies that we have in different departments.
So, in a way, we hope to solve any possible questions regarding how our company operates. So, if somebody wants to find information, theoretically, Conversational AI should be able to answer it. But we are far from there because it's not easy; it's not something that you can just switch on and expect it to work. But at least we are very optimistic that everything will go in the right direction. We hope that the UE will not block anything. We have to be...
Anyways, it’s a long discussion about AI. It's too early to say whether it's safe or not or if it’s good or not; there are many discussions surrounding that.
Kieran Gilmurray: It's interesting, just following on from your conversation a moment ago because AI just doesn’t work. You don’t just plug it in, and everything works immediately, but neither do people. You can't put a person into an organization and expect them to know everything from day one without training them or showing them or guiding them about what is right or wrong. I do wonder sometimes if we have too high expectations of technology and too low expectations of people.
Laurentiu Bogdan: Exactly, that's the essence of technology since I’ve been working in IT. Everybody expects that technology should work flawlessly from the beginning, which is far from reality, you know, it’s impossible to hit. And if we are going back to Conversational AI, it's exactly what you're saying. You have to invest time, knowledge, and know-how in order to make it better, let’s say.
So, of course, you have to be... It’s exactly like when you have a new employee; you have to teach them a lot of things. It will be easy if you are teaching only parts of the processes, it will become much more difficult if you want to make a megabot. I don't know what is the better word... you know, a mega-employee. Let's put it like that because now I'm thinking, I think it's impossible to replicate in a human what you can possibly do with a Conversational AI.
Imagine that if you were to put a megabot and an employee to the same task, with the objective of knowing everything about the company, everything, I'm pretty sure that the employee will not be able to do it. Instead, with conversational AI, maybe you can do it. There is some hope. That's what comes to my mind right now.
Kieran Gilmurray: I don't think there are enough hours in the day for a mega-person to actually exist. But that's the beautiful thing about a megabot. You teach it everything once, and it retains that knowledge forever. Whereas if you have to train 400 staff and then there's a new 400 staff, the amount of learning that has to happen, the amount of checking, and the amount of updating is phenomenal. But maybe, let's say we should do that competition.
Where do you see, Laurentiu, the future of generative AI in the pharma industry? Are there particular use cases, or what do you think is possible going forward in the future?
Laurentiu Bogdan: Oh, if you are looking at other scenes of the Generative AI, there are a lot of applications in the domain, like new molecule discovery. You can discover faster, you know, the time it takes to discover a molecule now takes almost ten years or more.
So, if you can reduce it to two months, it will have a big impact. And not only that, I think you can use it in all areas, more and less, it depends on the AI. I don't see any field that cannot be touched by Conversational AI.
Kieran Gilmurray: Oh, that’s a rather big statement because that's a lot of impact on a lot of roles.
Laurentiu Bogdan: Oh yeah, I know everybody talks now about losing jobs and so on. I think it will not happen. If you look back 100 years ago with the Industrial Revolution, in fact, we didn't lose jobs; we gained jobs. And I'm pretty sure that it will happen in the same way. I'm pretty sure that we will win jobs.
What's important to highlight is that I'm not talking about people here, I'm talking about jobs. And why I'm saying that is because some people can lose jobs if they cannot keep pace with the technology. I think it's very important to remain flexible and curious, to learn and adapt to new technology. Otherwise, of course, you will lose.
Kieran Gilmurray: But again, as an advocate like yourself, I believe that technology creates jobs. One hundred years ago, when technology put horses out of work, then it ended up with manual manufacturers, mechanics, and tire industries. It was a whole new era of innovation. People were able to move across great distances and create new value. And that's the bit I’m excited about Generative AI.
It's the alteration, and yes, some roles will change, but it’s the alteration of the makeup of roles. So, if I can have a megabot that answers every mundane question if I can get it to anticipate what I want on a daily basis, organize my schedule, organize a presentation, and free my mind to be my most creative and value-adding self, then I look forward to everything that Generative AI can bring. It's going to transform how we do things. I want that digital assistant today! Give it to me, and I'll start teaching it because that’s going to drive tremendous amounts of value.
Laurentiu Bogdan: Yeah, I think that's the direction. All of us will have, in one way or another, a digital assistant with us. But, it will just augment our capabilities, not improve them. I think that's the way we'll work in the near future and probably in the further future. I don't know exactly how the world will be different, but within the next 15 years, it will be like that.
Kieran Gilmurray: I’m hoping it does improve me.
Laurentiu Bogdan: No, I think it will help you. You know, I like this comparison, there are people who have a lot of ideas, and they are very intelligent, but they lack the ability to express themselves. This scenario can be very helpful for them. They will be better because they will be able to express their ideas, which can be very good. This is what I'm thinking about.
For example, I’m trying to write, I have a blog, and I don't have very good writing skills, but if I use ChatGPT, I am better. They are still my ideas, you know, but I have an assistant to make my text better. It's like, when you're writing a book, and you go to a publisher, and they will fix your book to sound better. It's almost the same thing, in fact.
Kieran Gilmurray: I like that. So, what's next, Laurentiu, for you, for Gelu, the assistant, and for Servier Pharma?
Laurentiu Bogdan: The next big step is to achieve the data integration with GPT, to have a chatbot, a Conversational AI that can answer most of the questions. I'm trying to be realistic. And then the next dream of mine is to give Gelu a voice. That's a big step because the majority of our staff is working from the car. They're travelling a lot. So, it will be even closer for them to use their phones to call Gelu in order to achieve their tasks using their Conversational AI assistant. That's the next big step we want to take. And, of course, while we have a clear vision of what we want to do, many other projects will likely appear around the journal because everything is popping up. You never know when a good idea will just pop out, and you say, "Now that's a good idea, let's use it!"
For example, in the last two weeks, we added to our Conversational AI assistant a method to schedule a parking lot rental from our office. You just have to say to Gelu, 'I want to reserve the parking lot,' and he says, 'Okay, it's done. You have this lot, that time, that’s fine.’ You can improve the knowledge and the know-how of Conversational AI very easily.
Kieran Gilmurray: That’s fascinating. I don’t know if people think like that, Laurentiu. You know, what I’ve heard today... you say it as it’s the most natural thing in the world, but the things we’ve heard today are not. You know, being an absolutely integral part of the business, not having one foot in and one foot out, understanding the business intimately and then giving them the technology through Conversational AI to answer questions they don’t even know they have.
To do amazing things like booking you a rental space, using the most natural thing possible, which is people’s voice. Well, giving Gelu a personality is not something you hear every day. And all of this, as you described, is within the same budget that you had ten years ago, and I love the way you gave the analogy that you get your salary and within sight of your salary, you innovate, and you spend it in different ways, but you’re still getting the same amount.
What you’re doing in IT at Servier Pharma is amazing. Whilst you’re not going to say it, I admire what you do, and I hope everybody else who’s listening to this podcast today gets inspired to suddenly invest in people and in great technology that allows them to impact a huge organization in the same way you and your IT departments are currently doing. I wish you every success in it and cannot wait to see what happens next!
Thank you so much, Laurentiu, for agreeing to talk to us in DRUID Talks today.
Laurentiu Bogdan: Thank you, Kieran. It was a nice conversation. I like to talk about that. Thank you very much!