Who are the Data Scientists working at Capgemini Invent?
Interview with Ahmed BESBES — Senior Data Scientist at Capgemini Invent
What is your background and favorite data science topic?
I have a background in software engineering and machine learning. Being a Telecom Paris alumni, I have thoroughly learnt about these technical topics as well as other subjects such as signal processing, computer security and networking.
I also spent my last year of studies at HEC Paris in a joint master’s program where I studied strategy, e-commerce and digital marketing. Although it was a fun program, this was not my thing. I was personally more excited by the technical challenges and the practical ways of solving them.
When it comes to data science, I have developed throughout the years an expertise in deep learning. I have trained and deployed many networks for several applications such as sentiment analysis, knee injury detection, automatic captioning, neural style transfer, etc. I like this field because it combines state of the art advanced theoretical concepts with experimental design and lots of hands-on coding as well many hacks, tips and tricks. Building a robust deep learning model is always a challenge, a fun one.
Tell us about a project you worked on at Capgemini Invent…
One of the most interesting projects I’ve worked on was designing a social media monitoring tool that analyzed union online communications in order to track the discussed subjects and prevent social events such as strikes. Our client was interested in this approach because it allowed him to have a global overview of what union was communicating over social media such as Facebook, Twitter and online blogs. We had the liberty of investigating all the data sources we wanted, to work in autonomy on AWS, and to use several visualization tools. Regarding the data science part, we designed many scrapers and we built models that tracked topics over time and scored their sentiment in order to highlight their critical nature. This tool was successful and later adapted to other use-cases such as marketing, e-reputation scoring and evaluation of citizen polls.
What do you enjoy in your work today, what motivates you?
Some people are relieved and satisfied when a project is done but I personally think that joy and excitement come out through the continuous building process. Building is a journey and this is what l like most about my job: the ability to craft things from the ground up, to collaborate with people on a technical issue and to see a product evolving and maturing over time, step by step. For me, this is the most exciting part and I have experimented this on some projects where we didn’t even have data first, but worked our way through the end to come up with a working solution.
I personally think that building, experimenting, trying and failing, getting feedbacks and retrying again is a stimulating process that seamlessly integrates with my daily routine.
And aside from work…?
I love technology and I’m big fan of open source and education. As a matter of fact, I spend a large part of my free time maintaining a data science blog, a github repo and a YouTube channel I created to talk about my open source AI projects. I love this!
Little fun fact (maybe): I built my own homemade GPU server to train deep learning models because AWS started to be very expensive. I named this server Zeus and I occasionally use it as a heating device when winter is around the corner 😊
