Maxime B.

Project context: Before the implementation of the UiPath solution, negotiators lost a lot of time generating all the contractual documents related to commercial offers. Now, they save several days on each sales campaign, thanks to the use of the RPA robot developed by DataPy.

Realization:

o Support for architecture choices involving an Unattended robot,

o Complete development of the project, in collaboration with a DataPy business analyst and project manager.

Technological approach: This project is carried out with UiPath RPA technology. A process is developed from a Framework on a studio and is integrated on VMs on the ATR side.

Inventory of flows and update of BODI (Business Objects Data Integration)

o Project context: The objective was to inventory all inter-application information flows within the company, plan their exchanges, and clear out those that no longer function or are unnecessary. Then, update each flow to make them compatible with the new version of BODI.

o Technologies approach: Manipulation of SAP Dataservices ETL BODI. Management of some SQL scripts to review. Excel for flow inventory and planning.

Creation of SQL scripts to integrate a new application into a data warehouse

o Project context: Omnium Finance has a data warehouse where each application stores its historical data via flows. Two new applications, which will eventually replace two old ones, need to be linked to this warehouse. To avoid modifying the entire warehouse structure, this project involves creating a mapping between the new data structure on the application side and the old structure on the warehouse side.

o Technologies approach: Excel for mapping different tables and columns and generating necessary scripts. SQL Server for testing and production. BODI for creating data flows to the warehouse.

Logistic regression P.O.C.

o Project context: The objective was to conduct a proof of concept to raise awareness within the company about Machine Learning technologies. The prototype is a notebook on Google Collab that determines whether an investor has a profile for purchasing real estate or an investment product.

o Technologies approach: SQL Server to retrieve data and store it in Excel files. Python for the algorithm and interface (pandas, numpy, scikit-learn) with Google Collab.

Refactoring of a Data pipeline for Skywise

o Project context: The automation of the creation of a dataset used as a source for a machine learning algorithm needs to be ensured through a data pipeline. This existing pipeline must be refactored and moved to a different environment (from an ETL to an IDE).

o Technologies approach: The existing pipeline is on an Airbus development platform called Skywise, using a tool similar to ETL called Code Workbook. The language used is Python with the Pyspark library.

Creation of a Data pipeline for statistical purposes

o Project context: Airbus wants to perform statistical analysis on a dataset concerning short-haul flights. During aircraft configuration, a modification tree has been populated for over 30 years. It has been decided to archive it and create a new tree that meets Airbus's future needs and can support a much larger production than 30 years ago. The client wants to analyze the existing data from a large dataset.

o Technologies approach: The pipeline is created on Airbus's Skywise tool: Code workbook (ETL) and a simple interface composed of the visual dataset and multiple applicable filters is made on Skywise: Slate (an equivalent of Microsoft Access). The language is Python with the pyspark framework. The interface part is developed using PostgreSQL and Javascript for managing real-time filters on the dataset and the ability to download it as is in CSV format.