Ali Fradi

Development of a data acquisition infrastructure accessible from multiple sources to support the study of urban investment opportunities. – Defined data sources, metrics, technologies, necessary features and offered services. – Wrote specs of ETL pipelines, defined scope and tools for data storage and acquisition. – Mentored developers and trainees about good practices for business-oriented data engineering. – Modularization of ML models according and data APIs to the subscriber sales business model.

Development of an exact solving method that integrates decision-makers’ preferences to optimize cleanrooms production in an Industry 4.0 environment. – Developed 2 linear program models for cleanrooms production planning. – Studied trade-offs between computational efficiency and formulation constraints. – Developed a dashboard for monitoring resource occupancy and allocated energy.

• Created ETL pipelines for NLP tasks, automated testing with GHA, and orchestrated with Airflow DAGs. • Multi-containerized ETL app using Docker, automated deployment with GHA. • Built financial sentiment models, Shiny apps for NLP, and a recommendation engine. • Automated scorecard models in PMML. • Introduced CRISP-DM for credit scoring. Integrated APIs into Shiny dashboards. • Participated in Power BI POC. keywords: R, Python, Power BI, SQL, statistics, PMML, Airflow, GHA, Docker

• Adopted CRISP-DM approach designed for industries to achieve the project. • Development of dashboards: automate calculations and EMEA reporting. • Modeling: prediction of sterility class of corn plants (hypothesis and data driven approach) to optimize the costs of detasseling plants and preserve the quality of the product. • General model prediction precision is about over 80 %. • Deployment of the model on local server. Keywords: R, RShiny, Python, mathematical modeling

Development, in collaboration with customers, of Shiny applications, test and validation of the application Sankey graph on two iterations and that of ANOVA since the first test. Observing fans’ interactions on Facebook radio pages in Tunisia to advertising posts through a customizable Sankey graph generated from the input data. Analysis of the effect and interaction between factors defined on the qualities of olive oil through automated statistical tests and ANOVA.

The mission aimed to increase sales of Shell lubricants in stations during refueling time Benchmarking and classification of necessary products. feasibility study and budget estimate for the realization of the project

Banking Finance and Machine Learning specialization

Linear algebra Mathematical analysis Engineering technological sciences : mechanical sciences Inorganic Chemistry