Shaping Your Data Science Strategy with Genentech's Youssef Idelcaid

Episode Thumbnail
00:00
00:00
This is a podcast episode titled, Shaping Your Data Science Strategy with Genentech's Youssef Idelcaid. The summary for this episode is: <p>Today’s guest rose through the ranks of retail giants L’Oreal and Levi Strauss before landing his current role shaping data science strategy at one of the world’s leading biotech companies. In this episode, we’re joined by Youssef Idelcaid, the senior director of data science for Commercial, Medical and Government Affairs at Genentech. Listen as he shares what surprised him most about his transition from retail to life sciences, his best advice for getting started with&nbsp;AI, and how he decides when to stay in-house and when to outsource data science projects. And don’t miss “Youssef in Context,” where we discuss his greatest career inspiration, a new title for your must-read list, and who’s invited to Youssef’s ultimate bolognese dinner.&nbsp; </p><p><br></p><p><br></p>
What surprised Youssef most when he transitioned from retail to life sciences
01:01 MIN
How the life sciences industry can begin to humanize AI for its users
00:55 MIN
Why transparency is critical for building trust and driving adoption
01:11 MIN
Youssef's guiding principle for driving transformation
01:04 MIN
Start, fail fast, iterate. Youssef's tips for getting started with AI.
01:31 MIN
Deciding where to focus in-house analytics resources and when to rely on outside partners
02:35 MIN
The many nuances of "context" in the healthcare space compared to other industries
01:42 MIN
Youssef's addition to your reading list
00:51 MIN

Youssef Idelcaid: There is no harm to dream big. We have to dream super big. We have to aspire to a smart, intelligent, maybe even superhuman future and put a vision and a roadmap for that.

Clay Hausmann: I'm Clay Hausmann, chief revenue officer of Aktana and host of Contextual Intelligence. Our guest today helped shape the data science strategy at one of the world's leading biotech companies, but he doesn't have your typical pharma resume. Instead, he came up through the world of retail, where he spent the last decade driving AI and digital projects at L'Oreal, and most recently, Levi Strauss. We're joined today by Youssef Idelcaid, the senior director of data science for commercial medical and government affairs at Genentech. How is his experience working in retail influencing the way he's using AI in his current role? Well, that's what we're going to find out. Youssef, welcome to the podcast, and thanks so much for joining us today.

Youssef Idelcaid: Thank you for having me.

Clay Hausmann: So, as I met mentioned in your bio there, you spent the last 13 years of your career before Genentech in retail. We tend to be very insular on this podcast. We are in the life sciences industry, we're life sciences people talking to other life sciences people, so this is quite thrilling to have somebody who's got the perspective much broader than our industry. So I want to ask you, especially for coming from retail, which is often held up as one of the models for AI, given its early adoption, at a high level can you just give us a sense of the progression of data science and AI and machine learning in the commercial model in retail during your time in that industry?

Youssef Idelcaid: Absolutely. For me, there are two big milestones, the 2000s with internet, there is for me before and after internet, and also there is a before and after iPhone with Steve Jobs, 2007, 2008, just because the evolution of AI/ML it's nothing but evolution in people's habits, the way we shop, the way we interact with internet and also with ourselves and each other. So the example I have for this evolution in retail is basically going from descriptive analysis to a predictive and prescriptive analysis. An example that comes to my mind is Target. You're probably familiar with the story of them predicting or sending marketing campaigns to a family. And the father came to Target saying, " Why are we sending this baby product to my daughter and she's in high school?" So it turned out that the daughter was pregnant. That was a little bit, I would say, the beginning of that experience in retail, having a static at way of analyzing things, all the retailers, having millions of receipts and sitting on that data on- premise warehouses and having to analyze them to create segments and ways of targeting clients based on what they're buying, pure transactional way of doing analytics. And with internet, and this is the after, basically we had pretty good sense of what is the potential of internet when it comes to e- com. So people still go into stores but we saw that a lot of businesses started selling things online. And that's created a new need. With that evolution in technology, it created the need of developing new ways of analyzing people, not making it pure transactional analysis, but making it link to behavior. And then we had people moving from internet to mobile. Everybody shops through different channels like social media and other marketplaces that are at the tip of our fingers. So basically, we saw much more personalized way of recommending and selling and also the need of customers has shifted into, " I need something now and I need something that fits perfectly my need." There is no trade off, there is no negotiation. So this is how I see the evolution of AI/ ML in the retail space, going from what I call static point of view to something very dynamic and personalized to satisfy the consumer.

Clay Hausmann: I have to say, that Target story, when it first came out, I found fascinating when it was first published, but many years later now that I am the father of a 15 year old daughter, I find it terrifying. So I have a very different perspective on that story as things have changed. So I'm curious then, it's mid 2021, Genentech comes calling to you about an opportunity and you decide that you want to make a shift from an industry like retail that's been deep into AI and machine learning for a decade or more and in the ways that we think about it. True data scientists would say, it goes back much further than that, but the way that we've talked about it as of late, and moved to an industry that's considered more of a late adopter for many good reasons, because there's a lot more at stake and there's a lot more that you're dealing with there and you decide, " You know what? This is a move I want to make." What was your thought process? What made you decide to make that shift and join Genentech?

Youssef Idelcaid: The role was just tailored for me. Like, this is exactly what I'm looking for. I will be hearing science all day long, but at the same time, I will be on the operation side and mix between my love for research, but also my love and passion for operations. That's one. Two was, I quickly saw the opportunity and the gaps, right? I had the impression that there is a lot to build and there is a huge opportunity to be part of, I would say, not a project or a program. It's for me, a revolution in the space of healthcare, because I have done also my due diligence and I realized that healthcare is not a space where digital transformation has gone far beyond compared to retail. I guess there is a number from McKinsey that says it's around 7% of healthcare industrials have gone digital versus 15 or 20% in retail. So the bottom line is, you put me in a field where everything is flat and the grass is already trimmed and beautiful, that's not for me. You put me in a space where it's pretty hard or difficult to navigate, but there is an opportunity to apply machine learning and AI in order to solve complex problems, this is where definitely I will be at the max of my capacity and I'm happy to contribute and build a strong organization that will support the company vision.

Clay Hausmann: So, I would imagine there have been things on both sides of the ledger. There have been some things that are, as you expected them to be, reinforced what you had hoped for when you made the move, and there are probably a couple of things that surprised you. What are some examples on both sides?

Youssef Idelcaid: There are gaps of course, that we're trying to bridge, but also I was really surprised in a good way, the amount of this workforce and resources, I think we can achieve a lot. And also, I'm super surprised how knowledgeable people are, especially on the CMG side. Yes, we are on the operation side, but people, they really understand the context and its just amazing how they can translate basically science into something completely tangible and operationalized and make sure that our patients are being served properly and customers in general. It doesn't go without some downsides where when you have these rock stars everywhere. It does create some silos. We have small groups. Everybody's launching initiatives on their sites, and I think the ideal state is to break those silos, so we need to remind everybody and remind ourselves that we work towards the same goal and purpose. And that's normal because it does confirm to me that there is a lot to do in terms of bringing also some innovation to the company and using machine learning and AI can be one of the biggest components to bridge the gap between the different groups and a unified pipeline of initiatives in this space. So, yeah.

Clay Hausmann: How about on the surprise side, maybe not about Genentech, but about the industry. Was there anything that, as you made this shift out of retail where you had been for a while into healthcare and life sciences, that maybe you didn't expect and you found, either it could be a very pleasant surprise that counters the perception, and I think you just mentioned it with the level of talent that exists in this discipline for an industry that's considered a little bit behind the curve in adoption, but is there anything that's surprised you as you've been part of the industry that you didn't expect coming in?

Youssef Idelcaid: Absolutely. I did not expect, for example, having a multi- cloud. That was a good surprise because I think, yes, retail is advanced, but when I joined Levi's in 2018, the only organization who had cloud instances were e- com people. The rest of the organization, for example, supply chain, operations, everybody was running on- prem still. Genentech, I came because it's the first biotech company in the world, and I did not expect having that multi- choice, I would say. If we trust the numbers in healthcare, everybody of course is saying that healthcare is not as much advanced, but resources wise, I think what I like also in healthcare, we do give ourselves the means and the tools to advance the purpose. So that was a good surprise for me, because most of the work has been done already, from tech stack standpoint and resources. The work or the hardest to come is shifting the mindset of people from legacy to working with data based approaches and tools. So you cannot, for example, for engagement, go and pause a tool or take technology to a rep who used to plan their businesses the old way. So there is a little bit of education to do, to the fact that machines are not replacing humans, but they're augmenting them. So good surprise for me. We have a lot of assets at Genentech and I'm very proud that the road ahead is still very long, but I think we are on the right path.

Clay Hausmann: Well, you're interesting, you just mentioned this notion of the humanization or the partnership between humans and AI and there's a lot of discussion about that in our industry. It's interesting because that same mental block doesn't seem to exist in a lot of consumer applications. So if Netflix recommends a different film to you or a different TV program, based on what you watched before, we don't have an issue with that quite often. And I know the statistics are high on how many of the views come from those personalized recommendations, how effective they are, or when Google Maps gives you a different route, that doesn't offend us in our ego and our driving ability. But in our industry, there is more distrust at the outset that needs to be overcome, whether that's between a sales professional and an application or an AI or a piece of machine learning that's supporting them in decision making. How do you think we overcome that? Why does that exist here more frequently? And maybe how are we more effective in normalizing this in an industry like ours?

Youssef Idelcaid: It has to do, I think, with the level of trust we build between the machine and user. With Alexa, for example, people never imagined they would need this kind of assistance at home to ask for playing music or ordering food or doing such things. And even with that, we had all these controversies with, "Alexa is listening to me and where is my data going?" But still people, because they get what they need as users, as customers, they start building this satisfaction, this trust. And also, the fact that all these providers on the market, in the internet industry in general pay attention to the privacy more and more, it kind of reassurers people, " Okay, it's fine if you have a pod of Google Assistant at home, and you ask that assistant to do some tasks for you. You can turn on and off things, you can activate, deactivate things, you are in charge, right?" Similarly, I think, in the healthcare industry, there are people out there who've been in the business since years. And they engage with healthcare professionals in a certain way. And if someone comes and all the sudden gives them an engine that engages on their behalf, or that gives them this kind of recommendations, first, they need to know where these recommendations are coming from. Second, they need to know, so what is the level of confidence you have to put in this kind of recommendations? And I think the fact that we trust Google Maps to bring us to the right right place or right destination it's because it has been experienced, it has been proven, right, over time. And I don't think they got it right the first time. It's the iterative approach that made Google Maps, what it is today. And we can still hear people comparing Google Maps to Apple Maps and having this kind of comparisons to say, " Hey, this is completely off." Humans criticizing the machine and so on and so forth. I think overall, building that trust and giving control to users, the perception of AI has been a tool to augment and support humans versus a tool to substitute to what humans can deliver. I guess it has to do also with this notion of feedback loop, as far as I can tell the machine what to do and override what the machine is doing, we're friends, right? So it's exactly that. But if I have zero control, and if I have the impression that I'm dealing with a black box, this is where we lose trust in AI and machine learning because, AI and machine learning is not the end goal, the end goal is the value it generates for me. What is it in machine learning and AI for me as a user, as a human, and how does that assist me to make my life easier? And it's not impossible. I have seen it at Levi, I have seen this at L'Oreal. People always start by resisting, but then, once you build the trust and you give them more control to be part of training the model. If you have people who are serving as subject matter experts, they generally trust the machine.

Clay Hausmann: I think it's very well said, honestly. And the way you describe it's not about AI machine and learning, is the endpoint. It's about the value that AI and machine learning generate, is very accurate. So at Genentech, you are leading and part of a major digital transformation effort, a lot of large life science companies right now, as they make this shift towards being more digitally oriented and as it accelerates, given the way it's changed behaviors coming out of COVID. That requires you as a data science lead to probably have to make trade offs, sort of balance out decisions and investments that will lead to innovation, maybe more long term versus more practical or pragmatic decisions about near term and hitting marks that you need to hit in order to maintain momentum. What is your decision making process yourself? How do you balance that? How do you make those trade offs?

Youssef Idelcaid: That is an excellent question. And we're in the middle of that as we have a lot of use cases, and I talked about me being positively surprised of the number of talents and initiatives in the company. So I have a guiding principle because I learned the hard way to be honest, there is no harm to dream big. We have to dream super big. We have to aspire to a smart, intelligent, maybe even superhuman future and put a vision and a roadmap for that. We have to paint that picture. However, there is also no harm to start small, because basically, in a company that is not a company that sells software, we don't sell AI solutions as a biotech company. This is not the core of our business. So when it comes to prioritization of what goes to this AI/ ML pipeline project or initiatives, I think the first thing to do is to start small and show success in a very niche area. I always say that we have to show the tip of the iceberg to the execs, and then we can show them the rest. To gain momentum, you have to start somewhere. The mistake we have done. I think when I was at L'Oreal for years, it's, we were having this empiric or sequential approach where let's work with IT and create databases for all our data, for all areas of research, right? Getting our master data correct, and we were so obsessed. And as a result, we haven't started transforming until very, very late. We did it and we succeeded. And today for me, L'Oreal is number one in beauty tech and all these kind of nice things, right? But to me, the first thing to do is just start. And the prioritization approach I have is, what are the key issues that are recurrent in the company, the things that keep occurring again and again, and for which we don't have solution like using AI/ML and data based approaches and things that are intense, things that are cost sensitive, things who are losing money, things who are losing competitive advantage. So that for me, a mapping we should be doing as leaders to really find that sweet spot, just start somewhere. And we don't have to wait until we get all our data in nice shape and form. And I really have this lean startup approach where start experiment, fail fast so I can iterate, otherwise, you'll never get started. I always have big plans for my team, for the organization when it comes to AI/ ML. I have also an approach that I learned called open innovation, which is basically, don't wait necessary until you scale up your internal capability. You can definitely go and seek help and expertise from the outside world. And sometimes innovation comes from outside, not necessary from inside. And you can trust me on that one because by experience... Sometimes, the same idea. If you pitch the idea internally, you may have some resistance. If you ask your friend to give the company a call and you pitch the idea from outside, your idea might get executed.

Clay Hausmann: I'm sorry. Let me probe on that a little bit, Youssef, about in- house teams versus outside partners. When Aktana essentially created the space of commercial and medical intelligence in life sciences 10 years ago, even seven years ago, most of our customers did not have robust data and analytics teams internally. We were playing that role quite often for them on the outside, but now, that's changed quite a bit. And even as you just described, the level of talent that you have internally on your team, as you think about projects and expertise, how do you decide where to focus in- house resources and what to rely on outside partners for? Is there a consistency or does it vary project to project, campaign to campaign?

Youssef Idelcaid: There is no good or bad answer to this question, but my experience shows that we need both. I like the example of Aktana because when I joined basically, my team started actually building some in- house capabilities and algorithms for next best action, for example, engagement. The problem is, when we had to start scaling things like for our partners in consumer engagement, we weren't staffed and we did not have the expertise like from software engineering and all this kind of talents to take that to the next level. It just seems a natural choice when you want to accelerate, sometimes it's great to go to the expert. Aktana is a software company, Genentech is a biotech company. So everyone has an expertise. There might be instances where the hybrid model can be the best for the companies where basically they build a hybrid model where both solutions coexist. You are still dependent on an external platform, but you are also able to co- develop things with that provider. And this is where I see a lot of opportunities for any company, not only in healthcare and biotech. And then there are areas where the expertise is a pure internal business, right? You better do it internally because you have all the resources and we are ready to go and you don't have any time constraints or any cost limitation to achieve that. I would say research topics or topics with a long range vision, I always tend to do internally. Things where I need a very fast execution and scale, I always advocate for working with outside world. I think it's just like when I have time and I don't have very timely sensitive deliverables or investment or ROI, so I take my time. And that's fine if you have, a core team who is developing or working things internally, because you can then explore things and no constraint. But if you know that, for example, in areas, if we don't do engagement, if we don't do next best action, if we don't have a solution that is up and running, other players in the markets will do it and that's a missed opportunity for the company. So that's why I think external partners do provide that competitive advantage for sure. Plus, the expertise, of course. I guess you have more data scientists that I have, then you have more software engineers that we have internally, which is completely okay because it's your expertise and our expertise is to make the best and the most accurate treatments for patients. That's what we do and you create the best technology that's going to help us do that. So I think that's my mindset at least, my philosophy on this.

Clay Hausmann: That's great. Well, one other area I want to ask you about, we're about to get started on a medical intelligence program together, and providing a 360 degree experience for HCPs across all their touch points with a company like Genentech, make sure that's a very consistent and valuable experience for them is critical. That's obviously what we're all seeking to provide right now. But at the same time, it's important to keep commercial and medical separate. So as not to pollute the medical council side with any commercial messaging or other influences. So how do you strike that balance? How are you looking at that for Genentech to make sure that the experience provided to the HCP is a full 360 one, but then there is enough division that's still important between those two parts of your business?

Youssef Idelcaid: I love this question. Yes. Our approach is the following, we learn from commercial way because the commercial system is very complex as well. It's too big, from data interactions, the data sources are huge. There is a complexity where we build the muscle in commercial and then for medical, I think, it's more of applying the technology, not the same data sources or not infusing outputs from commercial to influence medical because commercial is about promotional and medical is about science. Of course, both universes coexist when it comes to talking to the same HCPs, for example. We do believe that at the end of the course will be one- stop shop for both in order to have this integrated experience for our customers. But when it comes to designing and building things, there is a huge respect to compliance on the medical side, as well as on the commercial side. That's why, actually, we're trying to also include that compliance aspect to the initiative.

Clay Hausmann: Well, one thing I want to spend a little bit of time talking about is, you just wrote an article on healthcare recommender systems published on Medium, also up on your LinkedIn page. Everyone listening to this should follow Youssef on LinkedIn, you can find the article there, give it a little plug, but maybe, what motivated you to write the piece? You're five months into your experience at Genentech, but already you've written an article about healthcare recommender systems and the nuances, the terminology, a lot of the facets of it. What motivated you to put the piece up?

Youssef Idelcaid: To your point, I'm new in the industry and because I've been exposed to recommender systems in retail, I wanted to know what does a recommender system look like in healthcare? And the article I posted in Medium blog, it's nothing but a summary of that research. Is the first, the series of articles I will be publishing around the techniques themselves. It's going to be a little bit more technical for the upcoming ones, but it's a way for me to materialize the knowledge I'm building in this new industry.

Clay Hausmann: Well, one thing you mentioned in there is something that we talk a lot, because it's obviously the name in the podcast, Contextual Intelligence, and you talk about context.

Youssef Idelcaid: Yes.

Clay Hausmann: And you break it down into two key elements of contextual factors and multifactorial goal setting. Could you maybe describe for a minute, what are those elements and why are they so essential to a successful recommender system?

Youssef Idelcaid: The context is one of the three aspects of recommender systems. There is the item, the users and the context. The context can be broken down into two things. There are the contextual factors and the multifactorial goal setting. So what does that mean? Contextual factors can be seen from a lens of two angles. The first one is dynamic attributes and dynamic factors that are related to users and their activities. Example of dynamic attributes would be the best time to take fat soluble vitamins with dinner, for example, right? If you take an example of dynamic factors that would be more related to user or patients, and it's also dynamic and it has to do with the emotional state for a specific activity. So that's the first dimension or the first aspect of the context when it comes to recommender systems. On the other hand, multifactorial goal setting is taking into consideration domain specific criteria before recommending an item, as opposed to e- com where most preferred items are most likely to be recommended. It doesn't work this way. In healthcare, we have to completely shift gears because blood pressure lowering medicines are good for patients suffering from hypertension. These drugs can be super dangerous for the diabetes patients for example. That's why having this multi- goal setting for what's good for the user, it's very important. So this is the difference between for example, the retail and healthcare. Not that there is no context in retail, there is, but it's completely different and it's straight forward versus what we see in healthcare, because it can be having some serious consequences when it's not done properly. And machine learning, again, can help doing that with some sophisticated systems and algorithms.

Clay Hausmann: Well, Youssef, you've been very generous with your time. We've really enjoyed the conversation. I am going to try to steal a couple more minutes from you here for a section that we do at the end of every podcast episode, which is in this case, called Youssef in context.

Youssef Idelcaid: Of course.

Clay Hausmann: And I'm going to ask you a couple of questions about your interests, your background. And I'm going to start with a question that we always start with, which is, who has been an influence on your career that might surprise us?

Youssef Idelcaid: Steve Jobs. Definitely. I'm a huge fan. He inspired me even beyond the career. Design for me is very important. For example, when you develop a UI, it has to be easy to use for people. It doesn't have to be a super complex thing, design and the end goal of simplifying and providing this integrated experience for internal and external users, I think it's just beautiful. And again, I remember I never needed to have an iPhone, but every year I buy iPhone because it's just beautiful and it has a purpose behind. So, yeah. So Steve Jobs definitely, very inspiring.

Clay Hausmann: I don't think there can be an argument with that answer. Okay. So if money was not a factor, what career would you most like to pursue other than the one you're doing right now? So of course we know that this is your life's passion. You would do it for free.

Youssef Idelcaid: Yes.

Clay Hausmann: But other than what you're doing today, what career would you most like to pursue?

Youssef Idelcaid: I would definitely join one of the United Nation's agencies. I was born in a developing country, Morocco, North Africa, I think. I always been very sensitive to nutrition, water, all the essentials of life. And my parents came from mountains, south of Morocco, where they had to work and walk hours to get water. And being part of organizations like that, that's going to facilitate some of the access to this kind of things. And I think the UN is a good spot for that. I don't know, maybe one day. Maybe my experience with Genentech will lead me to work with WHO when I'm 50 or something, but don't tell my manager.

Clay Hausmann: Well, it's a very timely and appropriate answer. The perspective you just shared would be wonderful if that was shared globally with what's going on right now.

Youssef Idelcaid: Absolutely. Thank you so much. Thanks.

Clay Hausmann: At the other end of the spectrum, at the bottom of that list, what profession would you most not want to pursue no matter what it paid?

Youssef Idelcaid: Being a dentist. I'm traumatized. I just hate the idea of opening someone's mouth and pull... I don't want to do that. I don't want anyone to do that to me.

Clay Hausmann: That's good to know. So at one end of the spectrum, UN peacekeeper, at the other end of the spectrum, creator of pain as a dentist. I understand.

Youssef Idelcaid: Yeah, exactly.

Clay Hausmann: That makes sense. Okay. What is the best book, film or show you've enjoyed recently and why?

Youssef Idelcaid: Yeah, it's an old book, actually, 2014, it's called Superintelligence. I really highly recommend it. If someone wants to understand all the challenges and dangers of superhumans in AI, it's a good literature for the topic. I love what one of the critics said, " If superhumans or superhuman AI is getting bigger, it's the biggest event in our modern times, we have to make sure that it doesn't become the last event." And again, I liked your question about humanizing AI. And I think the trust is key. It's something super powerful, machine learning is super powerful, but humanizing that and creating this relationship between humans and machines, that's the key thing. How can we coexist with the different machines that are surrounding us?

Clay Hausmann: That would be wonderful. Okay. You're at a family gathering and your eight year old niece asks you what you do for a living, what do you tell her?

Youssef Idelcaid: I do have a eight year old niece. Did you know that?

Clay Hausmann: No, I did not.

Youssef Idelcaid: Did they tell you that?

Clay Hausmann: No. That's the question we ask everybody. You have an eight year old niece, yeah.

Youssef Idelcaid: Oh, my God. Yeah, she's Moroccan Belgian. She lives with my family in Belgium. I would just tell her I program computers and I hope she will understand what does that mean?

Clay Hausmann: Okay. So if she wants you to program her computer, you say, " Sure. Point me to it."

Youssef Idelcaid: Yeah sure. And then she'll ask me details and say, " Yeah, you can program computers to do some tasks, for example, you can program a computer to get you deliver a pizza or a coffee or something, or to get your jacket or skirt delivered to you or something like that." She will understand programming computers.

Clay Hausmann: All right, the last question. And I'm very interested in the answer to this one. So your ultimate dinner party for four, who is an attendance and what are you serving?

Youssef Idelcaid: Oh my God. My wife, my daughter, two years old and of course, me and our dog Safran. He's behaving, so I have no problem bringing him to dinner. Pasta bolognese, super, super Al Dente, Italian style with Parmesan.

Clay Hausmann: I love it.

Youssef Idelcaid: Very simple, efficient and we love pastas.

Clay Hausmann: I love that though, Youssef. At the same time, you dream big and you have this huge scope and way of thinking about things, but your answer to that question is perfect. There are also some great value in simple things in life too. Yeah.

Youssef Idelcaid: Absolutely.

Clay Hausmann: That's wonderful.

Youssef Idelcaid: Thank you so much.

Clay Hausmann: Youssef, thank you for your time. That was a wonderful conversation. I appreciate you coming on with us.

Youssef Idelcaid: Absolutely any time. Thank you so much for having me.

Clay Hausmann: That's it for this episode of Contextual Intelligence, I'm your host Clay Hausmann. You can find all our episodes on Apple Podcasts, Stitcher, Spotify, or wherever you get your podcasts. And please, leave us a review or a comment or a question or all of the above. So we can make sure that this podcast brings the proper context to your work. Thanks everybody for joining us.

DESCRIPTION

Today’s guest rose through the ranks of retail giants L’Oreal and Levi Strauss before landing his current role shaping data science strategy at one of the world’s leading biotech companies. In this episode, we’re joined by Youssef Idelcaid, the senior director of data science for Commercial, Medical and Government Affairs at Genentech. Listen as he shares what surprised him most about his transition from retail to life sciences, his best advice for getting started with AI, and how he decides when to stay in-house and when to outsource data science projects. And don’t miss “Youssef in Context,” where we discuss his greatest career inspiration, a new title for your must-read list, and who’s invited to Youssef’s ultimate bolognese dinner.