Ala Bayoudh

Data Architect / Engineer – Data Platform Projects (Leclerc Group, DPD) Led full lifecycle development on Google Cloud Platform, from planning and deployment to ongoing maintenance. Advised clients on technical decisions and provided training to optimize use of their data platforms. Delivered solutions for multiple entities within the Leclerc Group, including: Banque Edel, Coopelec, Isocel, Scalandes, Scacentre, Scafel, Scapalsace, Scamark, Socamil, Socara, Scaouest,Logilec, Leclerc-voyage. Tools & Technologies: BigQuery, GKE (Docker), Artifact Registry, Cloud Run, Cloud Functions, Dataflow, Compute Engine, Datastream, Composer (Airflow), Terraform, DBT, Looker, Azure DevOps, Cloud Build

I lead an AI for NLP course with Python, and I supervise engineering students on Data Science projects. Courses : NLP, LSTM/RNNs, LLM, RAG, Python

Data Consultant at EB Partners

EB Partners is an insurtech that supports insurance companies, mutual funds, provident institutions and brokers in France and abroad in their business and digital transformation. Data/business technical lead on a renewal tool, Data process migration from Access to Python, Customer relations and making specifications and decisions, production management and Sprint monitoring on JIRA

Early Signal is a project that aims to detect credit rating downgrades of large counterparties by S&P between 6 and 15 months in advance and to prioritize the cases for solvability analysis. My role: • To automate, and industrialize Python model with code optimization. • Integrating a new data source which is more exhaustive to improve the number of counterparts in the prediction files. • Optimization of data processing • Automation of data recovery (Manual to automatic) • Integration of a new financial database • Power BI visualization • Improved the number of counterparties in the predictions file from 1400 to 2500.

Building a list of skills and jobs using NLP techniques to help users better define their profiles. • Using regular expression to clean up job postings • Use of Spacy “PARTS OF SPEECH TAGGING” to generate Tags (Adj, verb, noun) • Created a rule-based algorithm to reduce skills (3 words).

Internship at FIS, the world's largest provider of banking technology and financial software, I worked on financial data and provided descriptive and predictive solutions to US managers that gave them more information on how transactions were being made and to extract specific characteristics, two approaches were used: -Unsupervised learning approach: using Kmeans and CAH clustering on financial data to create transaction clusters and determine each cluster characteristic. software used: (ML Azure Studio, Python jupyter Notebook) -Supervised learning approach: Create, test and evaluate machine learning algorithms for regression and classification, which helped create models to predict specific target variables using multiple independent variables. Algorithms used: decision tree, boosted decision tree, random forest, logistic regression, linear regression. software used: (ML Azure Studio, Python jupyter Notebook)

Web development with nodejs, Angular2

web development (angularjs, nodejs)