![\"Data](\"https:\/\/blog.atinternet.com\/wp-content\/uploads\/2019\/04\/Interview-Data-Scientist-AT-Internet-1024x678.jpeg\")
<\/figure>\n<\/div>\nDigital analytics is used to support a wide range of data science projects, and vice versa. Features built with machine learning and AI are also emerging in web analytics solutions \u2013 predictions, targeting, new services, etc. In a recent publication<\/a>, Gartner refers to augmented analytic<\/em>s as the main trend and priority for CDOs in 2020. We recently interviewed J\u00e9r\u00e9mie Bureau, data scientist and head of the data science team at AT Internet, about data science and analytics. Read on to learn more\u2026<\/p>\n There are many ways into the world of data science. Engineering schools and universities offer courses from master\u2019s to PhD level. Indeed, the demand for data scientists is so high that specialised private schools are starting to spring up. Personally, I studied applied mathematics at the University of Bordeaux, then went on to a doctorate in mathematics and statistics at the University of Toulouse. I wrote my thesis under the CIFRE agreement (French industrial convention on training through research) and worked as an R&D engineer for a startup during my three years of doctorate. My thesis was on the reliability of geolocation systems in an aeronautical context. I then worked in various professional fields such as health, employment and digital.<\/p>\n When working on issues that require data processing, before even going However, each site will have its own specific properties depending on The data science team needs to provide tools that target all our An example of this is a segmentation method to identify users who Data science is now able to provide descriptive, RFM segmentation is another use case <\/strong>\u2013 it is a clustering (segmentation) of customers according to their Firstly, the construction of a data science<\/em> roadmap in line with the Each member of the team must now be able to handle both modeling and industrialisation issues. The technologies and tools the team uses are very diverse: Python, R, Shiny, Scala, Spark, Elastic Search, Kibana, Snowflake, AWS, Kubernetes, Jenkins, Git, etc. The other key challenge is to ensure that the team\u2019s skill base progresses consistently for everyone. To do this, we work with platforms such as DataCamp<\/a> or Kaggle<\/a>. <\/p>\n It\u2019s important to always stay sharp and Digital analytics is used to support a wide range of data science projects, and vice versa. Features built with machine learning and AI are also emerging in web analytics solutions \u2013 predictions, targeting, new services, etc. In a recent publication, Gartner refers to augmented analytics as the main trend and priority for CDOs in 2020. […]<\/p>\n","protected":false},"author":1,"featured_media":87764,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[12033],"tags":[12765,21031,11603,5727,3229],"dealstore":[],"offerexpiration":[],"class_list":["post-87763","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-analytics","tag-analytics","tag-bureau","tag-data","tag-interview","tag-science"],"yoast_head":"\nHow do you become a data scientist?<\/strong><\/h2>\n
How does data science specifically apply to an analytics solution in SaaS mode?\ufeff<\/strong><\/h2>\n
\ninto predictive models or machine learning, we need to meet two requirements to
\nextract information that is actionable and value-added \u2013 firstly, collecting a
\nsufficient volume of data, and secondly, ensuring it is representative of the population
\nwe want to study. AT Internet<\/a>\u2019s huge advantage is
\nits variety of customer websites, making it possible to tick both boxes!<\/p>\n
\nits business sector. These differences can vary enormously from one sector to
\nanother \u2013 e-commerce sites, media, advertisers, banks, institutional sites etc.<\/p>\n
\ncustomers to help them optimise their marketing strategy. The tools, based on mathematical algorithms and models, must make it
\npossible to describe and predict the behaviour of Internet users.<\/p>\n
\npurchase the most, or alternately users who have a high probability of churn
\n(unsubscribing or not returning to a site). It is often a case of choosing
\nbetween a generic model with an acceptable performance on average across all
\ncustomer sites, or a specific model for similar sites.<\/p>\nHow and why is data science useful for web analysts today?<\/strong><\/h2>\n
\npredictive and even prescriptive tools to support analysts. There are numerous metrics to monitor and
\nunderstand to obtain useful information. It\u2019s also not rational to try to
\nfollow this vast number of metrics manually. One of the applications of machine learning to
\nsupport analysts is to offer an automatic
\nanomaly detection service<\/strong>. The goal is to capture unusual or suspicious
\nfluctuations in metrics over time. Our teams are currently working on analyses to
\nexplain the probable causes of these anomalies \u2013 e.g. if a bot passes over a
\nsite and causes a significant peak in traffic, an anomaly is detected on the
\nnumber of pages viewed. We aim to support the
\nanalyst in his or her investigative work by automatically exploring a set of
\ndimensions (source, device, browser, etc.). Our causality analysis module<\/a> shows that this
\nanomaly was caused by an abnormal increase in traffic on the direct traffic
\nsegment in Canada on the Chrome 55 version. This type of tool will enable the analyst to
\ncarry out an initial analysis and gain a better understanding of behaviour to
\nanticipate and implement the necessary actions or strategies.<\/p>\n![\"Anomaly](\"https:\/\/blog.atinternet.com\/wp-content\/uploads\/2019\/04\/Anomaly-Detection-Analytics-Suite-2-1024x339.jpg\")
\npurchasing habits to optimise a marketing strategy. Customer transactions
\nare analysed based on three criteria: Date of last purchase (Recency),
\nFrequency over a given period, Amount (cumulative over this period). Scoring methods are then used to create the customer segments, such as Stars<\/em> who buy a lot, and who have bought
\nrecently, or Thrifty dormants<\/em> who
\nhave a poor recency score. At AT Internet, we have decided to integrate an
\nautomatic RFM clustering feature \u2013 the idea is to use a turnkey analysis which
\nwill automatically adjust to the customer context and especially to seasonal
\nfluctuations. In addition, prediction elements are added and integrated into a set of
\nadapted graphs. Our teams are currently applying the same segmentation methodologies,
\nbut on metrics related to engagement rather than purchasing, to enable these
\nfeatures to be used on non-transactional sites.<\/p>\n![\"Clustering](\"https:\/\/blog.atinternet.com\/wp-content\/uploads\/2019\/04\/Clustering-RFM-Analytics-Suite-2-1024x320.jpg\")
What are the data science team\u2019s challenges at AT Internet?<\/strong><\/h2>\n
\nneeds of our users. Our priority is to be attentive and responsive. From an organisational point of view, our team
\nis now part of a high-level development environment. This requires the
\nimplementation of a workflow combining major R&D work, industrialisation
\nand continuous optimisation of our models.<\/p>\n![\"Data](\"https:\/\/blog.atinternet.com\/wp-content\/uploads\/2019\/04\/Data-Science-Workflow-AT-Internet.jpg\")
And to sum up\u2026<\/strong><\/h2>\n
\nattentive, with a passion for discovery and learning \u2013 \u201cData science is driven
\nby curiosity\u201d.<\/p>\n<\/div>\n\n","protected":false},"excerpt":{"rendered":"