Pharma Insights 🩺

Hi, everyone! Welcome to a new episode of Pharma Insights.

My name is Juliana Kreisel and
I will be one of your hosts for today.

As you may know,

Pharma Insights is a talk show created by Platforce

with the aim to connect, network and exchange ideas

from the pharma industry from all over the world.

Today we will explore how to use AI to enhance Commercial Excellence.

Our panel of experts will disclose
some key points regarding these new tools

such as how is AI changing our go-to market pharma,

and what will happen with omni channel.

This episode will not be only moderated by me,

I'm going to invite my colleague Stefan Repin to the stage.

Hello, Stef. How are you doing?

Hey! I'm doing great.

It's finally a good weather here in Budapest, where I'm at.

and we have the show today

and I'm really stoked about the guests we have.

It's like all superstars, you know?

So just let's introduce the other superstars.

Sure sure. Then our first guest

for today is Florent Edouard.
Hello! How are you doing today?

Hello! I'm good.

I'm in the Netherlands and it's not warm, as usual

Hahaha I love it!

So, guys, don't go there!

If you're looking for warm weather, then it's not your option.

Then I would like to introduce

our second guest for today Manuel Mitola.

Hello, Manuel! How are you doing?

Hello, Juliana. Hi, Stefan. Hi, everyone!

So all's good from my side.

I mean, actually I'm from Italy and today,
I don't want to brag, but the weather is great.

So... haha I'm very glad to be here.

Great! So, you heard:
go to Italy, don't go to Netherlands.

Yeah, for food and for weather.

There you go! And last but not least,
we have Marco Andre. Hello, Marco!

How are you doing today?

Hi, everyone! Don't come to Switzerland.
Summer hasn't arrived here yet.

I believe winter is coming here.

So, guys, we do not only bring you
information about the pharma industry

but we also bring weather reports from all over the world.

So, you know, just take our advice
and go to Italy ha ha ha

Ok, before we start I want to remind our audience

that we would love to hear from you

so please leave your comments and questions

below in the comment section

and we will bring it up here to the panel.

also I want to let you know that
we would love to hear from you

in terms of what topics do you want to hear after

so my friend, my colleague, and social media guru
Malén D'Urso will be on the comments.

So, she's going to drop the form, you can fill it out,

and we will plan our next topics regarding your thoughts.

So, without any further delay
I would love start this talk show

with this incredible panel.
Stef, how are you feeling today?

Yes, I'm feeling great!

In fact, we have so many topics to start with

and we have just discussed this,
before we started the webinar of ours,

that we want to go more into commercial AI

when it relates to GTM, go to market,

and since Marco was the one to bring it on,
maybe he should start it on haha

I'm thinking... I'm thinking, well,

you have AI now, right?

So I was reading this book about the go to market strategy

and I was thinking about AI at the same time.

How can AI help you with your go-to-market strategy?

But when it comes to pharma,

we just mentioned that for example

when you do a research on Google,
you don't get much outcome nowadays

but for example with tools like Perplexity

when you're looking for your ICP, ideal client persona,

it can give you like a way more, and a better outcome

than Google coule ever give you, you know?

and that, you know, if you have your ideal customer persona,

that's already the beginning of your go-to-market strategy.

But you have many other things to incorporate there

So my question is: what's a part of...

What's a basic part of the go-to-Market strategy for pharma

when it comes to, well, particularly pharma?

What are the building blocks
nowadays when it comes to AI?

Anyone can take the question.

I just, so when I need go-to-market,

I know this conversation me and Manuel had last week.

Literally the best way to think about go-to-market is

where the market is going?

And where the market is going,
if you think about HCPs nowadays

we have reports that they're using
these technologies to get answers

much faster and much better.
Patients the same.

So if we think about the immediacy, and the speed

and how they're using these tools to get those answers

It's unlikely that will go to other sources

so I always give this example

a lot of us having the industry HCP portals,

which need to be gated,
which people need to put a password

need to look for the information,

need to download some information, and scan for it.

Will HCPs in six months, in one year actually be using that?

It's like any of us when we started to use

desktop, and then we pivoted to mobile

How many hours a day do we
spend in mobile versus desktop?

So that's the first thing I think we should look at.

It's how we adapt our strategy based on

where the behavior of our cost...
in this case our HCP and our patient is going.

I think I would even take it earlier than that,

I think I'm very lucky in the world of commercial excellence

in a sense that you know when joined Grünenthal in 2018,

I was able to completely shape
the whole commercial excellence organization.

The only thing I got was the people

like there were people, there was no structure,
no mission, no vision.

and one of the things we have established very early in comx,

is that we need to be the ones

who are going to drive the creation of the insights.

in the organization, specifically around the go-to-Market model

so right now for instance I got a beautiful asset

that we want to launch

in a specific therapeutic area

and the key question is

which physician do we need to talk to?

because they have the patients
that will benefit from this treatment

and that's going to generate

the right Revenue stream for the company.

And this is really for me the starting point

where AI can help us.

And for instance if you take
just an example from the US

you got the claims data, which are

billions of rows of prescription by individual in the US,

which is data that you can access.

The only way to manipulate that data,

to identify where those patients

are being treated along their patient journey

is to use AI models

and once you got that,
you can then know where those patients

are being treated and you can

interact with the customers at that point in time

where they will be facing a patient.

And that's opening completely new grounds

in terms of possibility to be

to be accurate, to be relevant

with the physician and with the patient

so really for me it starts you know at the beginning

and then all along the commercial journey

AI is going to reshape the way we do commercial excellence.

Yeah, if I may add, Florent,

a couple of items let's say

came to my mind where you and Marco were speaking.

The first one is related to the the speed

because I think that when it comes with AI,

this is an element that

is going to be changed massively.

We were speaking before

how much time do you need to do that?

how much time do you need to get an Insight from the data?

so on, so I think speed also from our side,

let's say from the pharmaceutical industry,

if we were thinking about a go-to marketing strategy

with a certain lifespan in the past

now we are I think shortening it.

so that's one key element.

The other key element that it is let's say close

also to my own experience.

I have been working into
the Omni Channel marketing field

for more than eight years and

and I think we have been always saying

you know, right channel,

right customer, right target, right timing,

and so on, we have been always hearing

this kind of thing I don't know how many times.

I think that

I mean to be honest we never went there

we never really went there but I think

that AI maybe will be and is the

thing that taking the data from multiple sources,

having the power of the

instilling Insight from this multisource data

maybe will be finally

able to enable a true Omni channel so

I would also say that maybe

the Omni Channel as is today

will also stop existing somehow

it will be a completely different and new thing so

in my mind I have these two things top of my mind.

Yeah, I agree. I'm totally with you

and one on the speed

like you know

it's not impossible to imagine tomorrow

that we will be able to monitor
in real time the social media

to detect adverse event and
address them in real time as well.

Today if you tell that to

the people in the company,

everyone's going to tell you you're completely nuts.

I mean, that's going to cost us a fortune,

and we don't even know if that data is accurate,

but with good AI models,

we are not that far away from being able to monitor that

and from the Omni Channel standpoint, it's the same thing.

The good old days we were segmenting customer

on potential and adoption

and they had to be in the right box

high potential and high adoption or you had to move them there

That was very rudimentary,

but that was better than nothing.

Today with the tools that we have

with the capture of all the omni channel interactions,

we can segment customers on 200 and plus dimensions

and delivering really One to One content interactions

and think that they really need
in the format that they like.

So, yes, I'm totally with you.

Especially when they don't have the GDPR haha

Yeah, sorry, Marco.

And just on Omni Channel thinking about go to market

we might be facing a future
in which there's no Omni Channel

we might be facing a future in which

each one of us interacts through only one

you know form of Jarvis or assistant

in which we will go through all of the other things

but actually we might be moving towards convergence.

and I think a lot of this is guess work,

because a lot of, as you said Florent,

there's the insights

and everything is moving at such a speed

but one thing is for sure

either on the applied AI side or generative AI side

our industry is going to have to be much faster

and how we create but also then

send to Market, that
I have absolutely no doubt.

Sorry, you can go, Stef.

No, I saw that we have a question.

So I guess we all know Claude.

Claude Weddington, another influencer in the pharma space.

and Claude has a question, right?

So, Marco Andre on the spot, right?

Marco Andre, Novartis appears to have the most significant

patent cliff to overcome

considering those present in the conversation.

This is mainly due to the loss of exclusivity for Entresto

and the settlement-related patent cliff for Humira,

how active is the Novartis organization

leveraging AI for medical affairs excellence

in particular to combat the upcoming generics competition?

Yeah, we were discussing generics recently with Claude

and I think next year is going to be like

in two years there's going to be

like a ton of...

patents are coming... are expiring so

a ton of generics will flood the market so

this is I guess a question related to that.

So, Marco, the stage is yours.

Claude, good to see you.
It's always a good excuse to go to Portugal.

There is definitely better weather than Switzerland. Always.

So, listen. Obviously I cannot comment directly on your question.

Because there are some things

I absolutely can't comment on in public.

I can... I'm sorry. I know that's not the answer you were looking for.

I think what I can tell you is

how I am thinking about generative AI

and what I'm doing in Novartis

and it has to do a bit with what Florent and Manuel

what we're talking about is

if we think about this

theme of generative AI 18 months ago

probably none of us

knew what it was. I didn't know for sure.

We always see this graph of

generative AI that chatGPT

took two months to get to 100 million users

but that's compared with other apps,

right, with like Uber

but the one stat I found

is that it took 37 years for electricity
to get to half of US households

and smartphones it took 21 years.

ChatGPT took 10 months.

So I don't say this in a fluffy way that
we need to be kind towards each other is

this is a massive earthquake.

My friend says this is the biggest

systemic technological change that
any generation has gone through

So I think

even before we start thinking about

patent cliffs, and next developments

in medical, in applications in medical affairs

or commercial Excellence or R&D

there's a work to be done

and that's what I've been trying to do is

showing this to people,

showing it's not the implication
of a video being done very quickly

it's not so much that the video is done so quickly.

It is the way you interact,

as Florent was mentioning earlier,

the way you interact with your agency,
with your partners

and what you expect of them,
and of your own teams

changes, so that's the work

I've been doing in Novartis.

First, talking with everyone that I can

to make them at the most senior level

to show them that this is not a fad

it's not going to go away and
it's going to have massive impact

in whatever they do

so I know I'm not answering the question.

You're gonna understand why

but I wanted to give you a bit of my approach.

Then we have another question.

This is why people don't like pharma.

Guys, remember hahaha

ha ha ha

No one has said that you know
having a patent cliff is a good problem

because that means you have a lot of revenue for now

so at least, you know, you should say that.

ha ha ha

And then we have another question:

Can you enlight us more on the best
AI models that can be applied and how?

Who wants to tackle that?

Maybe I can.

Yeah, please!

Just to give my two cents here,

I am not let's say a developer

but I saw a study from the Stanford University

and also again the Stanford University has a website

where they are continuously testing the model

so I think the answer is

it depends on the task that you have to do

and basically you have different performances

for different, for different models.

I saw very good performances

generally speaking for Llama,

for example, I saw how to add

with the GPT and so on but it...

when you dive deep into the thing that

you have to do with the model then it is

then it is...

a different scenario, so it depends let's say

in short.

I would agree. I think it's difficult to

to say, you know, there is a winner per category

but depending on the problem that you have,

you will use a generative AI model

like GPT or Dall-e or whatever

If it's a compliance issue,

you may use another type of models

uh more NLP or you know image models

If you're talking about

you need to manipulate a lot of sales and market data

to understand where your market opportunities are

it's going to be sales predictive models.

I think there are really a lor of models.

The only thing for me is really

it depends also on how much money you have

and how many people

which means if you're a mid pharma company like us,

we don't have the luxury to have

tons of data scientists

manipulating and creating models and stuff like that.

So for instance when we go next best action,

we use models that, from partners

like Salesforce or others

because those guys can invest

in the tech and develop the model and stuff

and we can't afford that

I imagine that when you're on Novartis for instance

it's easier for you to build your own

personal model on your own data set

and be able to really protect your IP in that field.

I think it doesn't come cheap

the good models I think.

Can I add just on that, Florient

I think it might be good but I think

some of the decisions we need to make

and that's even outside the industry is

it might be the best

but it might mean not the most effective

so if I think about Bloomberg when they did the LLM

they were trying to protect their data

and I applaud them because
they were trying to do that very early

and they spent $10 million dollars.

Nowadays you're going to chatGPT,
you get a better answer.

I think a lot of the decisions we need to make is

exactly what we buy

like what we use, chatGPT, Co-pilot

what we build all on our own but then how we partner

how we get the best of

industry specific application, CRM for instance

and get our data

and I think the decisions to do one
or the other or the other

will actually make or break

our companies in terms of going to market,
so i just wanted to add that.

Yeah, thank you.

And another thing to consider just to close the loop also

is the fact that we can also use models

in combination, in parallel

and that's something that we can consider to..

Ok, we have another comment

from Jennifer, she says: Marco, good to see you.

haha I'm going to overlay the chat

so it's easier for us to see everything.

I don't know why... Oh there you go!

It says: Great to join this webinar today.

I would love to hear how the panel is seeing or planning for

use of gen AI in the context of

of Precision, Health. Oh sorry Precision Health

How we can identify patients with disease
or with uncontrolled disease?

with the goal of getting optimal medicine to them

in time or earlier

and reducing healthcare cost?

So maybe that's a little bit

what I was alluding to when I was explaining

how we chose which physicians we go to
based on the patient that they have

what you can do

and that's US only unfortunately,

we don't have that level of data yet in Europe.

So we need to find a way to get to the proxy level

but it's really literally

to understand okay

I've got my treatment

the patient who is probably going to be the most responsive

to that treatment has those characteristics

has been seeing those specialist,

has been seeing 10 doctors in the last five years,

has been on those type of treatments,

and has those you know characteristics

in terms of social demographics

with that you can really target

the right patients through the right HCP.

How do we get to that in Europe?

I don't know yet honestly.

It's a very complicated data set

that needs to be of high quality

what we can tell though is to give the doctors

descriptive of who those ideal patients are

so that they can detect them in their own patient pool.

But it has to be through a proxy

through the HCP.

Ok.

Just to add to that quickly

A lot of what we think

generative AI broad we think is
only new tech applied to new data.

It's also new tech applied to old data.

And I think that's why we've been seing

incredible progress.

I think Precision medicine is one of the most

incredible areas where we're going to have more impact because

the new tech applied to old data is

not every big company has an army of data scientists

and now you can actually go

human what would be humanly impossible

you can apply the data and derive insights

from something that you've been having right now

So again, as Manuel you were saying,

theme of speed

and what you can identify in those things

it's something that generative AI enables.

Thank you for that. We have another question from Claude.

Jennifer, brilliant question.

I'll be curious to learn

how Novartis is leveraging data42

in the aspect to train AI models for

enhanced commercial excellence.

I think, Marco, that goes to you.

It's like Senate hearing of Novartis.

ha ha ha

I'm going to have to give you the same

the same answer, Claude.

There are some things I can't,
I can't obviously answer.

to you or the public, or in private. I'm sorry.

Haha! No, it's perfectly fine.

Ok...

Which amendment is that, Marco?

Is it second amendment that you cannot incriminate yourself?

I'm not in US! I'm not in US!

ha ha

Yeah

So, now, Stef. Do we have a followup?

We shared the link that Manuel was talking about.

Stef was correct

that was Manuel for the Stanford University.

But that's okay.

Stef, can you?

do you have another question or another comment?

Yeah, so

in fact we do have a question from

Mr Adeel Siddiqui and his question is

how can AI...?

how AI can be implemented to do

to do segmentation and targeting for HCPs?

What is the role of commercial excellence

in segmentation targeting of HCP?

Again, how AI can Implement

to do segmentation and targeting of HCPs?

What is the role of commercial excellence

in segmentation and targeting of HCPs?

So well, for me, I mean

that's personal right?

Commercial Excellence

is there to give the rule in terms of

segmentation and targeting.

I mean if you want a neutral ground between medical, marketing, sales,

market access... That's in commercial excellence's land.

So that's the first thing.

The second thing is you need someone to process all that data

in a fair way and to decide
"okay we need to apply this type of things"

and we need to get it implemented into the system.

Again you are now coming back to

the two dimensions before,
and the more than 200 dimensions that we have

in the salesforce customer data platform today

it takes a lot of

steering

and that's an understatement

it takes a lot of steering to get

to each customer being profiled

willingly on 200 Dimensions

with high quality of data

and to maintain the data quality over time

so that your segmentation and targeting remains valid

so I really think it's on us.

For this one for me it's really in the commercial excellence and

my personal advice is

if someone comes to you saying

oh I got that beautiful algorithm that is a blackbox

and that's going to speed out a segmentation and targeting,

you press delete

and you send them to hell

because you need to own your segmentation and targeting.

Otherwise

you don't know your customers

and you don't know what you're doing,
and you're just wasting your money.

I think one thing if I can add, Florent, just on that

I think it will be increasing...

again going back to the behaviours,

so if doctors as an example physicians are using

declaring their intentions

and for instance a GPT

so for instance there's this tonic app that I saw a presentation of

and they're actually almost a companion

and if those companies are not
at certain point making that data available

how are we going to know how to segment them

if we don't know their intentions

so I think for us to your point is

what we do with the data but

how are we going to have access to that data

to know what are their behaviors and their intentions

that's going to be a challenge for us.

but I think a little bit to what, sorry,

I think it was Marcus

I think Marcus was saying we can only control

thanks to GDPR

only the customer that interacts willingly with us

and their data, that that's the only thing we can really control.

We have to make sure we do a good management of that

but in this case we can know

for the other ones I agree with you

there's no way we will know what's happening on GPT.

And from my side I fully agree with you

and I mean you have to own

let's say your segmentation and targeting, absolutely.

Maybe one thing that AI can

can help there

it's like it's an additional tool

and I think that

when it comes to segmentation and targeting

first of all it will be one let's say building block

for building the hyper personalization

also not only in terms of content creation but also

in terms of touch point flow generation

and another thing is that AI is very good in

finding let's say pattern, trends

that maybe from a human analyst standpoint

we are not able to see at first glance so

I see it is an additional tool

for the people working in Commercial Excellence.

Absolutely fine, I mean, I absolutely agree with you.

I would never say

Okay, push the button and

that's your segmentation and targeting

done by the algorithm.

I mean, if you come back historically

you know, why did we do it in two dimensions before?

because most people

live in two dimensions, you know.

Very few people like Stefan can live in three dimensions

with All Star Wars behind them, right?

I mean we usually live in two dimensions.

And that's why we simplified it this way

but now with the tools, the system

can create 300 500 600 segments

and manipulate them at the same time

doing differentiated campaigns to all those segments

and no one has to know exactly

which customers in which segment doesn't matter anymore.

Yeah.

Guys, we have another question.

Hi, everyone. I got one question.

Beyond lead scoring can AI predict...?
Let me show the question.

Can AI predict the closability of a lead

allowing sales rep to prioritize their efforts?

I can take that one.

So one of the industries I always admired

and as a reference from in Pharma is the B2B

software as a service and I think they perfected this.

Three years ago,

Claude is going to be happy because this I can disclose.

Three years ago when I was in Sandos,

we started doing even before AI was a thing

a pilot project with the consent of HCPs

to record their conversations

and it was actually incredible because

the amount of data you could get from that,

the training that you could do,

the prediction of Adverse Events for instance

I mean Adverse Events is something quite serious, quite serious

that sometimes it doesn't even get

raised because it gets lost, right?

So all of the possibility that you can do that

what are the most asked questions?

how good it would be, we were testing some things at the time,

to go from... get the top 10 asked questions about a certain drug

in a certain market

and have it updated on a daily basis so

I think that is something that B2B has been doing it.

for a lot of time.

Obviously there's a lot of compliance, legal,

consents that we need to iron out

but I think we will all need to get used to

that having our calls recorded internally and externally

will just become the default

and from that I think there is incredible possibility

So I'm a big big fan of that.

Guys, anything? Someone else wants to add something?

Or we move to the next question? Manuel? Florent?

Yeah yeah. I totally align with Marco.

Yeah!

I would like to...

I think Marco is right when it comes to

AI and SaaS, in SaaS you really have to

use lead scoring models but

there is a very big problem with SaaS that

these lead score scoring models do not work.

And they do not work in any industry.

because they give you

they give you some predictions based on

let's say I'm watching this webinar with yo

and it says "okay I'll give you a score"

I don't know 15, right,

and the score of 15 means then

maybe I'll click on your website,
on novartis.com, right? I'll go there.

And that will give me another five points.

And 20 points

I'm supposed to talk to a Medical Rep or sales rep

if you talk about business

and this is, you know, I might not be in the market to but this

The same comes for phama, like HCP can...

You have a personalized journey you know

Now I have a meeting with a medrep

and I'm going to go and research

through perplexity AI which we mentioned today

a certain drug for example

or for maybe a rare disease

that I'm encountering with my patiens but

the AI scoring model means that

by doing these actions which

basically me as a company I can score you know

I don't know what's happening back there.

It's a complete Black Box

but I know that you interacted with these points

and then I'm going to give you...

You have this touch points, I'm going to
give you a score, and based on that

I'm going to presume that you are in the market for

for the drug or you're ready to

start getting samples and tests.

I think it's... We're not yet there

because a lot of it is

in the dark funnel which we cannot track

and...

I think Omni Channel solves that issue

where you don't have maybe in particular

that much pressure on the HCP to take a decision

in your favor because

your goal is to influence them

and not to score a sale right away, right?

so your goal is more to basically be

interface and help them out, make a

a good decision, make an informed decision

and not you know prescribe your drug for

for, you know, no matter if it's good for the patient or not.

I mean that's my 5 cents.

Well so

I think we are in, I agree with you,

not to score a sale, but we are in to make sure

that a patient that can benefit from our medicine

will have a chance to get onto that medicine

if he's the appropriate patient, right?

so that's something that you can know

like, you know, I used to work in respiratory, right?

When you launch your respiratory biologics

you know that only

a certain number of respiratory specialists

are going to adopt those biologics

the other ones will adopt them

maybe in seven years later

when it's going to be very mainstream on the market

with a lot of experience.

By detecting those early adopters

being able to train them on the product,

making them understand the science

and giving them all the element that they need

to detect the right patient

we allow that patient

that patient cohort

to go on the product

so that the physician can realize
if that's a good therapeutic option

and that's what I would call

"closing the deal"

That's when we have done our job correctly.

Yeah, I like that because it doesn't need to be linear.

That's the point. It doesn't mean that lead scoring

is for you to make a sale

but lead scoring is for instance

if you get the top asked question, if certainly a client

is asking a lot of questions about your product,

let's take an example of B2B

is one either because he doesn't know

second because he's interested

So your job is not to just make a sale
of something he doesn't have information on

is to get him or her the information that he needs.

So that he can do a better job

and then just to add on Florent's point

there's an incredible potential

for AI with patient endurance

incredible if you think about some of these things

a lot of these treatments don't work

because of patient endurance

all the things that you see in Nvidia nurse

if I could have someone remind me every day

some form of GPT

of how am I doing, monitoring with my Oura ring

how am I doing with...

imagine if it's a condition related to that

and reminding me I need to take my meds

I mean that is incredible for endurance so

good point on raising it, Florent,

because we shouldn't forget that part.

Thank you. We have a one question that was on the chat.

but it got...

We have a lot of questions actually, not one.

But the first question that we have is from Jennifer.

She said: one more question

or perhaps controversial topic

at the end of the day we need health decision makers

to accept and changed behavior

hcps, patients and policy markers,

Health Systems, governments, payers

how does industry

Pharma, data, aggregators, developers, etc

most efficiently support

fund and influence stakeholders?

Maybe I can start with this one because I was reading it

in the chat before and I think it is something really

interesting, it reminded me, let's say

some years ago when we were approaching

the digital channels let's say

first sales call, virtual sales calls,

emails, and so on

and I think when it comes to this

adoption of Technology, of course you have

have a curve

with the the early adopter and

from the early adopter to those who are let's say

very late into the adoption

from my perspective

and that's my personal opinion

I think that the

let's say, the AI Revolution will

will take place will happen and it will be not led

only by the Pharma industry so the HCP will be in the middle of

a product that he or she use into the clinical practice that maybe

are AI augmented so more and more people

from let's say consumer to professional will be

more and more used to work with the AI

and I think the key point for us as pharma industry

is to be ready

and to, let's say, be ready very very soon

in order to let's say

do not frustrate the early adopter that are of course

those who are accelerating an adoption of

a technology so I think that's

something important

of course there are all the other stakeholders

all around I think some of them will be faster,

some of the other,

especially I think about, for example,

the policy makers maybe they will be a a little bit slower

but I think it really...

It really depends stakeholder by stakeholder

in Europe I think we are in a good shape

let's say with the AI act and so on

So if I compare our environment to other countries

I think at least there is some

effort, some attention that is going

into the direction of giving a framework, giving a structure

but yeah that's what I...

what I think about the question, Jennifer.

Thank you for reasoning

because it's a very good topic.

Yeah, I think just on this topic I believe

science-based transparency

in phrma, that this is

what will make the real difference

I think the era of

the marketing and the bullshit

and the sales rep carpet bombing is over

and right now because the patients and the hcps will have access

to neutral AI models

that will give them briefings on the products,

on the disease, on the symptoms,

on absolutely everything

If you are not transparent, if you are not honest,

you will not be heard and that's a huge difference

versus the past and

products that are not actually

differentiated based on science

and make it transparent, those ones will fail.

You can launch it, you can spend billions

on marketing that stuff will not take off.

And that's a game changer for pharma.

We have another question for Florent.

Let me put it here.

It says: on the topic of unstructured data

how do you handle unstructured data enhancement

to enable quality recommendations

on whatever data sets you may have?

clinical, KOL/DOL, portal data or data that you may buy?

Wow! So... hahaha

In terms of unstructured data,

we've got multiple types of unstructured data

so we have unstructured data

like the historical data of the company

I would call them unstructured
because they are all types and formats.

so those ones we are going to try to clean

document we are trying funnily enough

to use AI to be able to analyze them and tag them

using a very broad taxonomy

so that we can at least know what we have in house.

Then we've got, you know,

all the unstructured data that is coming back to our systems,

like you know the physician voice

voice of the customer, voice of the employee,

you know all that,

so for those ones as I said we don't have the means

and the capabilities to deploy

and create our own models so we're going to use

Salesforce, sentiment analysis, next best action,

and tools like that

to be able to analyze all that unstructured data

and get insights coming out of it

and then there is the unstructured data

for me the best example

is Twitter or X or you know

and those ones right now we don't touch

because otherwise we would be totally overloaded

you need I think to bite one size you know

after the other, just like not everything at the same time.

So we know it's suboptimal

but right now we're ready to try to

control and structure data internal

the external that is generated in our channels

and then we go to the external unstructured.

Okay.

Then we have another question.

I told you guys we had quite a few questions.

It says: how do you describe the implementation?

And do Pharma usually relies on

external agencies or it can be implemented inhouse?

I can take that.

Yeah, please!

I don't think there's a one size fits all in this case.

I think it needs to be

not only case by case but the situation

I think there there's three things that are really important is

decisions you need to make is right now

you need to rescale your entire organization,

like absolutely

in Pharma and outside of Pharma

so how are you going to rescale and going about it?

Some people may decide

we go every... We put everything on the agencies.

Right, okay.

Some others we bring everything in house

but if you bring everything in house

you need to make sure you have the right tools,

you need to build them, or buy them, or borrow them

and you need to train your people

and I think the training piece is

even before we start thinking about these things,

if we, and I put myself in that category,

if we don't really understand the impact
that AI can have in the value chain

we can't be making those decisions.

So Florent knows I'm a big fan of
the work he's been doing in Grünenthal,

because I feel as

it's not even as an industry,

we started focusing on the tech, on the data

in the use cases.

Guess what?

They're saying that generative is a flop

And it's not! It's like we didn't give the proper time

to train people.

When we moved from a calculator to excel,

we had to be trained and I think now is the same.

It's just a tool

so that's the thing, even before making those decisions,

we need to make people aware

of the implications of those decisions based on the tech that is used.

Those are my two cents.

And I fully agree, Marco,

I think you touched a very very important topic

that is the training and the re-skilling part

personally let's say I don't think

it would be let's say

for a pharma company

it would be the right move maybe from the very beginning to

do everything in house because of course

it is a fast-paced environment

so I think there are a lot of

players outside that are doing just the thing

and they are doing it very good so I think

a mix could be as usual

a good, let's say, a good solution

but pharma companies cannot

ignore the reskill need

because otherwise, and then starting from the business

because if we say "okay, AI is a an IT thing"

That's something that is

becoming a flop because people i

not able to use it, so you give an Excel

a software to a child

and of course it's not working.

It's something that is important.

To come back on the first part of the question,

even that's going to make me look like a Maverick,

I want to say that the implementation is

a fantastic moment.

It's super exciting.

It's not that often in my career

that I could go to a group and tell them

"there's a wave, 100 ft size wave,

that is going to land on us

and I need 20,30 people

who are going to be crazy enough to

with zero budget and no external support

are going to surf this wave together

and I had 20 to 30 ends raised

in the room, because it's super exciting

and if you take it, you know, with the learning eyes

and let's try and let's fail so

we have decided for instance that at the same time

we are going to implement agile at scale

so we are launching 15 Pilots

that we are running in an agile mode

so that we can kill very fast what should not survive

and we can move to the next pilot

but the people in it are doing it on top of their time

They are strapping the heart to find money

to fund the pilots

and they do it because they think

it's a game changer, so I would say

it's complicated but it's so fun

I would encourage everyone to try it.

Well, Carolina Quezada is saying

so she was responding to what Marco said

and she said: it's important to have fun

So, there you go, Florent haha

but I know Carolina

and yeah, it's true I mean

if you don't, I mean seriously,

we are doing a fantastic job.

We try to bring back the health to people and stuff like that.

That's a great mission.

But we are also human, we need to have fun!

If we don't have fun, we need to change

We go to the entertainment industry,
or something like that.

You know? If it has to be boring from Monday to Friday,

then just don't do it.

And this is a way to create fun for all.

And if I can just

I've asked to put on the chat

there's actually an article from last week

which exactly says that

so we might sound a bit crazy yeah you know

have fun and play and everything

but actually you can see that managers

that have actually played with generative AI

and show people like they show examples to their people

the engagement and adoption

is 350% higher

so take a look at that because now
there's data that actually proves

that the playful thing is not having fun

is actually way more effective.

Nice.

We have another question.

It says: so we're talking about commercial setting

do we need to be looking at proof of concept to use in trials 1st?

So first I don't know

but that's, so it's really interesting, you know.

When you go, so,

we are normally very siloed between R&D

and the people who are doing the clinical trials

and the people that are in commercial.

Like, you know, we are not supposed to be

working too much together

and in fact what we see

that those guys in R&D are moving very very fast

and they have been using

without anyone knowing it AI for years

to select and create molecules,

to select patients for the clinical trials,
and to do that type of stuff

they are actually doing it

but because they are R&D people

and not commercial people, they don't make a big fuss about it

that's a big difference so we can be their first maker.

Interesting.

Do you guys, Manuel, Marco, do you have anything else to add?

No, from my side, I mean I fully agree with Florent

I do not have big experience on R&D.

Oh, please!

I got a question, it's not related to this,

I'd like to dwindle the topic a little bit so

we have

so, you know, we talk a lot about AI,

and how it helps us with a go-to-market strategies

so go-to-market strategy you have

part one is targeting the right doctors

and the patients

and the other part is like field enablement you know

so the question is how can AI help us with field enablement?

I'll let anyone start with this.

I can take a pass at it.

It is related with the question we were

having earlier on...

on lead scoring.

We will all, regardless of our function, need to up our game

because access to information

the barries to enter are super low

so there's this story I always tell of this of this lady

she went through 17 doctors

three years trying to diagnose her son

after she was getting desperate

she took all of the notes, the last exam, and in half an hour ChatGPT

she got diagnosed

so patients are going to become super patients,

HCPs are going to become super HCPs

which means in terms of field

will need to become super reps

we will need to know what they are looking for

what are their questions,

they're going to come way more informed so

what we were discussing earlier again

to your point, Stefan, is about

about the lead scoring, is not so much as we will need to be

ready to have interactions with them

because they will know way more

than they could know before

and AI will help us

not only generate documentation but gather insights,

gather notes, allow us to be more present,
we always forget that.

People think I'm a romantic saying
that AI will make us more human.

but if I'm in a conversation with Manuel

and if I'm focusing on taking notes and

getting action items, everything,
I'm not fully present with him.

and I think that's how AI can help

tremendously on sales engagement.

Another thing to build on what you were saying, Marco,

I'm thinking of a practical example of

when it comes to field

so if you are for example entering

a new therapeutic area with a new compound

in an area that it is unknown for you

for the company of course

I think that AI can

play a crucial role in

let's say taking the data we all know

we can take the data from IQVIA for example

or from other service provide

like that

but I think the competitive advantage
will be also the usage that we do

of the AI to do for example a field sizing,

a field force planning

done in a very laser focused way

because we all know that of course

let's say unfortunately or not

the access of the sales rep to the Physician
is decreasing every year

so of course we have to

maximize as much as possible

the sales force if we want to leverage on this channel

and I think AI can really

help a lot as a tool.

It's just... I'm reacting also on
the message from Chiara in the chat

The other day I was watching a movie

I can't recall unfortunately the movie

but there was someone, one of the people

people in the movie that was describing symptoms

and I was hearing that

I went on to chat GPT

and I was... you know what? okay those are the symptoms

what is a disease?

and chatgpt gave me the right disease,
like same in the movie

So it's not a real life setting and stuff like that

but it shows how powerful those systems can be

and how much they are going to influence the patient

because then

before it was "okay, Dr Google is telling me that"

the doctors rapidly said "Dr Google is a donkey"

"so don't believe what Dr donkey says"

but tomorrow it's going to be Dr ChatGPT and by the way

Dr ChatGPT graduated as

a generalist in the UK

so don't believe it's a donkey

or you know UK donkey haha

but I think it's really really important we take that into account

in our go to market strategies

the patients will be educated by

the AIs tomorrow.

I definitely believe that.

We were having a discussion with my husband
about something, I don't remember what it was

but I was feeling a little bit sick

and the first thing that he told me was "Don't Google it"

because, you know, usually you describe your symptoms

and there you go haha

but now that I know that chatgpt can actually

gave me what I have, then I have a new doctor haha

No, I'm kidding haha

I know that, and that's scary because

then you're wondering

okay those systems have been educated using data sets

what was it that data set?

if you're starting to touch the health of the people

did someone pollute

the data set intentionally?

so that the system is going to give
the wrong answer to the patient

I mean, it can.

when you think the manipulation
that can happen in that field

it's like literally very scary

so we need to be very careful and
we need to educate our role I think

as trusted HealthCare Partners

is to educate the people on

how to use the system and
what are the limits of those systems.

Well, yeah...

Please.

No, no, sorry, good ahead.

I know, like Chiara said: Indeed, Florent.
Medical protocols will also improve

as well thanks to AI inquiring power and all patience insights gathered.

Yeah, totally.

And, Florent, you made me think about a question

that I received during another webinar

and I will not make any names

because of the question,
but basically the question I received

was that: is there a way to make sure that the custom GPT

is going to recommend our drug
instead of the competitor's one

after the patient asked for the symptoms?

I receive this question haha

ha ha yeah

Even that, but that's a reality, I mean a lot of people are saying

"You have SEO, right?"

You're going to start having some model
of large language model optimization.

That's going to be a whole new...

So we also need to be ready for those disciplines,

new disciplines to emerge, because, yes, of course...

I just want to quickly comment if I can just on Jennifer's

how do we bridge R&D and commercial

and I would love to hear it

the biggest concern I have right now
for the world and for the industry

is tech people

are not speaking enough to business

and business people are not being tech enough

I think we are in a very polarized world

I'm not going to lecture us on that,
we know different reasons for that

but my biggest concern is that

if we in business don't try to become a bit more techy

if the techies don't start become a bit more business

if R&D doesn't go closer to commercial

if medical affairs doesn't go closer

to any other function

I think we're going to fail on bridging this.

So that bridging word

triggered a lot for me

and that's what we need to do,
because I think if we do that

everyone is going to benefit

Pick the most skeptical person in your organization

go and have a coffee with them,

then we'll have, in complete denial of AI,

just have a conversation with them.

Thank you. We are two minutes away from ending this webinar.

So, guys,

do you have any closing comments for our audience?

I know you guys have a really really packed agenda today so

I want to be respectful of your time.

So, who wants to go first?

Manuel?

I can just...

Florent?

Sorry, maybe I will just...

because that's a link, I will just build on Marco's comment there.

Totally, our biggest enemy

is not external, it's not our competition,

it's not our biggest enemy is

our internal Silo mentality

where I take care of my stuff and
I end it over to someone else.

Those are the walls we need to break

to create a united data platform

on which we can build our models

across function for the benefits of patients and HCPs

and that means, to Marco's point,

business , IT, everyone talking and working together.

Otherwise, we won't be successful

and that expands in my mind also to partners

like openAI, like Veeva, like IQVIA, like all others,

because they play a role in our ecosystem.

and that's for me the only way

we're going to maximize the positive impact of the AI wave.

So sorry, Manuel,

I just jumped in because

that was connected to what Marco was saying.

No, no problem! I mean

From my side, let's say, I think

first of all I fully agree with you both

for both points, from my side

I would like to close with

something that I'm also doing personally

that it is to learn also from other Industries

because of course our customers are also customers

from other kind of products

in way less regulated environment

meaning that they can have more speed, more agility

and so on, so from our side,

from a pharma standpoint

we could have the luxury let's say to learn

from those who already tested in a safe environment

so, learn also from the other industry.

Thank you! Marco, something to leave or...?

No, the last thing is just, as as Florent was saying,

we're all different. we're all from competitors

and we're all here being very open

with the fact that we all have the same challenges

and I think we need to do that way more
if we want the industry to go forward.

So, thank you for inviting us.

And I really appreciate the open and candid conversation.

Thank you. Thank you, everyone. Thank you, audience.

And just to let you guys be

very calm and don't worry about me

I'm not going to use ChatGPT as my doctor

I will go to an actual doctor ha ha

So thank you all for being here.

Thank you to our audience.
We hope you have a great rest of the week.

And I will see you all on our next "Pharma Insights"

So it's gonna be on the 27th of June.

27th of June, guys,

write it up, and very soon
we're going to be starting a community

We're going to be giving you more information about that.

Thank you so much for being with us today.

I know it's hard.

The European Championship is starting.

A lot of games, you know.

Good luck to everyone!

Thank you very much! Thanks!

If you want good food and good weather, go to Italy.

I will see you all in our next "Pharma Insights".

Bye bye!