Thibaut is an Associate Professor in Outbreak Analytics at the London School of Hygiene and Tropical / Imperial College London (LSHTM / ICL)
He is also a Senior Data Scientist for the World Health Organization (WHO).
Jeff: Thanks for taking time out of your busy day to answer some questions relating to your data science process. But let’s start with the basics – can you provide a high level overview of what you do?
Thibaut: I am a researcher in outbreak response analytics, with a background in biostatistics, population genetics, and R programming. My research focuses on developing new methodologies and tools for understanding how infectious diseases spread, and how we can control them. In other words, I use data science to help respond to infectious disease threats, such as COVID-19 or Ebola.
I have also created the R Epidemics Consortium (RECON), a not-for-profit organization aiming to develop free data science resources for responding to health emergencies.
Jeff: Well, I’m sure you have been incredibly busy for the past year, and I thank you for all the work you have done to help limit the spread of these terrible diseases.
Can you provide a bit more detail on your work in ‘outbreak analytics’?
Thibaut: I focus on emergency outbreak response, in which analytics directly inform public health decision-making. Beyond infectious disease modelling techniques used in academia, I focus on the development of operational analysis tools, including reproducible and auditable data cleaning, interactive data visualisation tools, and automated report generation systems.
I also have on the ground experience of outbreak response. Since February 2020, I have been working full-time on COVID-19, setting up data pipelines for the modeling group at LSHTM and the WHO, and developing statistical approaches for informing the response in the UK, as a member of SPI-M, and internationally with WHO.
Back in 2019, before COVID-19, I spent 6 months in North Kivu, Democratuc Republic of the Congo, for the response to the largest Ebola outbreak this country has faced. I set up the analytics pipelines within the emergency operation centre, first in Béni and then Goma, for informing the leadership of the response on various aspects of the outbreak in real-time.
Jeff: OK, let’s talk about Data Science Process – You are Team Lead Certified. How has DSPA’s Team Lead course been helpful to you?
Thibaut: The course was extremely useful. It provided a concise yet complete overview of collaboration frameworks and data science lifecycles, which I think are very much under-utilized in academia and in public health emergencies. My only regret is not having taken it sooner.
Jeff: What pain points do you think the knowledge gained via the course help minimize?
Thibaut: It is easy to get lost amongst the many data science lifecycles and collaboration frameworks. Online documentation abounds but is often confusing as it tends to mix frameworks, workflows, and development philosophies. This course does a very good job of distinguishing these elements and explaining their respective roles and interplays. And it does so efficiently: I managed to go through the course content in about 4 weeks while working full time on COVID-19. I could have imagined this taking quite a lot longer.
Jeff: How has the course made a difference in how your team will work together?
Thibaut: This course was an eye-opener. Not only will it help me improve the way I manage my team, it also prompted me to create a new workflow and collaboration framework for outbreak analytics. After taking the course, I am drafting a playbook on how our outbreak analytics teams should use an integrated lifecycle and collaboration framework. The framework will leverage the key concepts of Data Driven Scrum and will work with the short timelines which are inherent in our emergency work.
Jeff: Can you say some more about the playbook you are creating?
Thibaut: For emergency outbreak analytics work, our team of data scientists and stakeholders comes together from a wide variety of locations and backgrounds. Since the team is created on short notice and needs to start working quickly to help with the infectious outbreak, we need to have a process that is easy to use and quick to learn. However, we currently work without a well defined process, basically people doing their best to generate and then use data-driven insights. So, the playbook will help ensure the extended team is working effectively together on the highest priority tasks, and that the insights generated are useful in helping to minimize the impact of the outbreak.
Jeff: What was your favorite part of the actual course?
Thibaut: Most content was clearly laid out, and follow-up discussions were very interesting. But the best part of the course was probably the weekly one-on-one interactions with the lead trainer, which were extremely useful, and helped me develop a workflow and collaboration framework that can be adapted to my line of work.
Jeff: What has surprised you most about the course?
Thibaut: The format was surprisingly easy to follow. Short lectures, followed by interesting discussions amongst the lecturers, made the material very easy to digest.
Jeff: Maybe one last question – what would you say to others considering taking the DSPA’s Team Lead course?
Thibaut: I wish I could go back in time and follow this course a few years ago. This course will be useful to a diverse audience: from people who are vaguely familiar with collaboration frameworks such as Scrum and want to have a more complete view of the field, to absolute beginners who have never even heard of the Agile manifesto.
I would recommend this course to any data science team lead, be it in the private sector, in governments, or indeed in academia, where such concepts are usually alien.
Jeff: Thanks again for taking the time to answer these questions!
To learn more, explore training via our Data Science Team Lead course.
In addition, you can register to get our monthly update of our blog posts, and/or jump in and learn more through these articles: