Mohamed ACHELIF

For 3 years, I participated in undertaking an ambitious technological project to decommission (decommissioning) a large monolith, MyDKT. Within the User Analytics team, I actively contributed to the development of Decathlon's new Data Platform while supporting Data Owners in the decommissioning process of tools from the old ecosystem. As a Senior Data Engineer with DevOps skills, I also participated in the design of various architectures: - Data Exchange Pipelines with key partners such as Valiuz, HBond, Epsilon, etc. - Data Streaming from the Data Lake to different Data Domains. - Migration of legacy ETL architectures on Talend to more modern solutions on Airflow and Databricks. I actively evangelized the importance of automated testing and best practices within the team, thereby improving the quality and reliability of deliverables. Stack: Scala, Python pyspark, SQL, DeltaLake Airflow Databricks, EMR S3, Glue, Redshift, DBT Crypto Github Actions, Mill, CI/CD, Kafka Kubernetes, Podman Collibra Slack for alerting and monitoring

During 3 years, I made a long technological journey on a project of reinternalization of the eCommerce of Kering while going towards omicanality (Eagle Program). Within the BI/Data Analytics team, I started to set up a backend for BI: external Hive tables (the data being on S3), then I boosted the platform with a Hive LLAP to get sub-second responses and finally I added Druid cubes for real time requesting. While we were beginning to notice the immaturity of Druid from an integration point of view with Hive, its complex implementation in cluster mode, and that it only has a native connector for Microstrategy and Superset, Athena emerged as a solution more adapted to our need: it is serverless, compatible with standard SQL (thanks to Presto), it has JDBC and ODBC drivers, uses Hive DDLs and all this at a lower cost! As a Senior Data Engineer with Devops skills, I participated in the development of different applications, each with a unique responsibility: A Scala application responsible for ingesting into the Data Lake An application for the transformation of data from Raw to Formatted, Aggregated and Usage layers An application dedicated to launching Glue Crawlers and updating tables/views on Athena and also responsible for updating data sources on Tableau Server. An application for the infra as a code A functional application for the Viz I used NiFi for most of the collections because of its simplicity and the increase in productivity to create new workflows. The orchestration and scheduling of jobs were also provided by NiFi. For the CI/CD, I set up, for each application: A Docker image with the app's prerequisites. A Jenkins pipeline to compile, test, push QA code reports on Sonar, assemble and push bundles on S3 Release management with sbt-dynver For monitoring and alerts, I set up a Pipeline Monitor on Slack based on Breadcrumb Trail. Stack : #Scala #Spark #Hive #LLAP #EMR #S3 #Glue #Athena #AWS #Jenkins #SBT #CI/CD #AgilePM #Kanban

As a Lead Data Engineer in the BSC/DAT unit, I participated in the development of a Data Lake for Compliance on HDP (HDFS for storage, Spark for processing, Hive for requesting). I also created an out-of-the-box Scala module called DataLogger in order to centralize and homogenize all application logs. DataLogger is a SLF4J compatible lib, so it can be integrated in any application running on JVM (especially Spark jobs). DataLogger is embedded in a Stack I created called KNEK: Kafka NiFi Elasticsearch Kibana. Stack : #Scala #Spark #HDP #Hive #CI/CD #Ansible #Kafka #Elasticsearch #Kibana #Agilité #Scrum

At PriceMinister, I worked on the Canopy project which aimed at dismantling the monolithic application of the PriceMinister.fr website by adopting the Microservices architecture and the DDD model (Domain Driven Development). Considering my senior profile on Scala, I brought a great contribution to the project by implementing the new features and developing 2 Scala libraries: ReactiveConfig (presented at Scala.IO in 2017) and Asynchronous StopWatch. I also made a PoC on PWA (Progressive Web App). Stack : #Scala #REST #Microservices #ReactiveProgramming #PlayFramework #Akka #PWA #DDD

At CarBoatMedia, I worked on a project to decommission a monolith application called Adams by adopting the Microservices architecture. In this context, I have participated in the development of 2 applications to manage the listings of the 2 websites LaCentrale.fr and AnnoncesDuBateau.fr: a Backend application called Rasalghul and a Frontend application called Valinor. Stack : #AWS #DynamoDB #Lambda #SQS #Scala #REST #Microservices #ReactiveProgramming #PlayFramework #Akka #Angular #Typescript #JS

I designed and developed a community site that connects people who have commercial space available with others who are looking for a place to sell. In other words, it was an AirB&B-like website for merchants. Stack: XP, Play 2.3, Scala, MongoDB, ReactiveMongo, AWS, JSON, JQuery, CoffeScript, Bootstrap, Less, Git

Agile development of CoachClub brand websites such as: www.coachclub.com, us.coachclub.com, domyos.coachclub.com and hundreds of co-branding partners. Application of the Scrum methodology within the IT team.

1. Development of the application entitled "Hermes" for StudioCanal. Description: The main purpose of the application is to identify and invoice the contracts contracts. All the sales of the movies of the companies Studiocanal and Canal+ Image UK are present in Hermes. Tasks: Design, development, unit and integration tests. Technical environments: RAD, Java/J2EE, JavaScript, Struts + Taglib + Tiles, WebSphere, Oracle, Invoice and pdf document generation with DocJar __________________ 2. Development of the "Technical Extranet" application for StudioCanal: the application is an intranet application accessible to professionals (rights buyer, theatre operator, journalist, equipment manager, etc.). These professionals order technical and promotional material for the movies for which they have an access right. Tasks: Design with Toad Data Modeler, writing of detailed specifications, development, unit and integration tests.

1. Development of an application called FAF (Filtering Against Fraud). Description: confidential application. Tasks: Design with Rational Rose, writing detailed specifications, development, unit and integration tests. Technical environments: Java/J2EE, Struts, JOnAS, $U, Shell, Oracle __________________ 2. Development of an application called DPF (Fraud Detection and Prevention). Description: confidential application. Tasks: Design with Rational Rose, generation of C++ code and design documents using Rational Rose. Technical environments: C++, ProC, $U, Shell, Oracle

Development of an application called DirSync (Directories Synchronisation) for British Telecom. Description: The goal of the internship was to design an application framework that would do : 1. Extracting data from data sources such as: flat files, databases, ... 2. Processing of these data: controls, calculations, ... 3. Loading: feeding or updating the company's directory with the data that has just been processed. Tasks: design and specifications, development, unit and integration tests. Technical environments: Java/J2EE, Struts, Tomcat, OpenLDAP, Directory One Server.