After Brandon Rufino completed his master’s in clinical engineering at the Institute of Biomedical Engineering in 2021, he leveraged skillsets in coding, machine learning, and biomedical engineering to design adaptable platforms that helps answer business questions at Sanofi.
What kind of work do you do as a data scientist at Sanofi?
As a data scientist on the Digital Data team at Sanofi, my job is to make sense of large amounts of data and develop analytics to help answer critical business questions.
Data in its raw form is not informative to a business, so here’s where analytics and summarization tools come in. These tools can generate insights about a business problem or provide opportunities that eludes the human eye on a quick glance. We have a duty as an industry to more effectively leverage data, digital and technology to change the practice of medicine by discovering and delivering transformative treatments to patients. The goal here is to use artificial intelligence to screen through large sets of data, and then quickly and accurately predict future hypothesis. We call it prescriptive analytics.
The Digital team has ambitions to build the leading digital healthcare platform. This platform consists of integrated data, digital and technology solutions to develop and deliver medicines faster, better engage HCPs and pharmacists, and help patients improve their health. Our use cases support our various businesses, and to name a few: use more data (RWD, genome etc.) and artificial intelligence to accelerate drug discovery and optimize our clinical trials; scale and personalize HCP engagement through multiple digital channels; refine our e-commerce solutions; optimize our marketing spend though data; and digitize our manufacturing facilities. If you look at the data in each of their silos, there are unique operational, security, and monetary considerations associated with each of them. For example, in a randomized clinical trial, we must think about factors like: how can we identify the correct population of people to participate in a clinical trial? How can we encrypt their data so it is secure? Each step requires an intricate communication between the data sources and the business teams consuming the insight. Ultimately, we want to design AI platforms that can pull data from each of their silos and generate insight to a user’s business question across the whole pharma chain.
What are some use case examples?
I work within the Research and Development vertical in Digital Data. Here, the team focuses on pre-clinical and clinical applications that accelerate making life-saving miracles for the human population.
Some examples of use cases within Digital Data Research and Development:
Drug discovery: A business unit wants to find out which mRNA vaccine should be tested in the lab for efficacy, prior to testing on cells. We can use active learning in artificial intelligence to tackle that, based on previously existing data. This would save us time and money in the research phase.
Clinical trials: Real World Evidence is making our clinical trials more efficient by reducing the number of patients that have to be enrolled, or enabling them to provide us with their data from a distance using digital biomarkers.
In the other areas of the pharma chain in Digital Data we have use cases like:
Logistics and transportation: We now have several nearly fully automated, remote-monitored, insight-generating digital factories and are improving our supply chain using predictive AI.
Commercial and marketing: We are deploying direct-to-consumer ecommerce platforms and have deployed a new and improved customer relationship management system that improves our interactions with healthcare professionals.
What are some of the questions you are helping to answer for the clinical team?
One of the common questions we try to answer is, which segment of the population responds better or worse to a particular drug? This is specifically tailored to Randomized Clinical Trials, where we are trying to cross-reference Real World Evidence and use previously existing Randomized Clinical Trials data sources to observe how a drug may perform in different population segments. All of this serves to increase the probability of success to have a drug with high efficacy and safety.
Another common question we try to answer is, how can we take a drug currently in market and find new indications currently not treated with the drug that may benefit from its use. This positioning of our portfolio allows our drugs to provide potentially life-changing treatments to a wider population.
There are a lot of existing data on previous drugs that came out of Sanofi’s pipeline. We are examining whether we can use these massive troves of data to help us train AI algorithms that can answer the previously mentioned problems.
Why should Sanofi focus on Data and Artificial Intelligence?
While the healthcare industry is slow to adapt, tech giants are moving into the space quickly and start-ups are increasingly disrupting the well-established beliefs, processes and business models. Whether we’re shortening development times for our R&D teams or providing new digital interfaces for patients worldwide, it’s clear that embracing digital is key for us to chase the miracles of science to improve people’s lives. Sanofi must catch up to its peers in certain areas, and leapfrog them in others, in order to remain relevant. Digital will be a key catalyst in this company-wide modernization.
In the first phase of our transformation, we are modernizing our foundation while also creating value for the business. Our modernization is focused on cloud migration, scalable, democratized and AI-driven data solutions, and a more efficient infrastructure. In parallel, we are already creating value for the business across R&D, Manufacturing & Supply, and Commercial through priority initiatives agreed upon by the ExCom.
In the second phase of our transformation, Digital will accelerate the value creation across our GBUs and GFs, create completely new digital businesses, and move us to an AI-driven decision-making organization.
The advantage here would be a more efficient, time saving, and objective systems that can rapidly answer business questions and provide future insights.