Hamza HACHANI

- Raw data scraping and storage. Technical stack: Python, Airflow, Minio - Implementation of a real-time data streaming library based on pyspark and kafka. Technical stack: Python, Pyspark, Spark Structured Streaming (batch mode and continuous mode)(api python), kafka, akhq, Kubernetes, Docker - Integration of this library in a FastApi project for the calculation needs of the API. Technical stack: Python, Pyspark, Spark Structured Streaming (batch mode and continuous mode)(api python), kafka, akhq, FastApi, PostrgeSQL, Kubernetes, Docker

Implementation for payments of an anti-fraud software based on business rules and machine learning models: - Development of Big Data jobs for real-time data processing - Implementation in production of a data lab allowing the production of machine learning models and the study of data (validation of the architecture file and then implementation until production) - Implementation of the SAML protocol as a backend service. - Automation of the deployment of technical components produced internally and those provided by the editor - Recipe automation - Definition of the scope of the successive versions to be deployed, writing of user stories, negotiation of the specification of the communication interface - Project management

As part of a project to create a central Data Hub,I am in charge of studies related to the DROOLS rules engine as well as the communication between BIG DATA technologies (Spark, Kafka,...) and an IBM proprietary environment. Management of a call for offers for the deployment in the company's CA TS data hub of a data encryption tool

As a Data engineer, I analyze the billing data produced by the "data access" service of mobile customers in order to verify the relevance of the billing systems (establish correlations between the use and the billing of the Mobile Data service ) .

As a research and development engineer, I set up a "Fast Data" architecture prototype for the recovery, storage and visualization of "End Device" data in the Thingpark solution, more particularly Thingpark Wireless and Thingpark Actility company store. The goal is to make data ingestion fault-resilient and scalable.The prototype is based on Apache Kafka technology.