Nihal Akel

• Loading back-office data into the BUDI database on a monthly basis.

• Developing SAS code to improve results and processing time.

• Documenting the process and results.

• Implementing business rules, data dictionary, and data mapping.

• Automating loading using SAS macro programming and performing non-regression testing.

• Migrating UAT and production to Gitlab using batch scripting to have a centralized tool for processing business requests.

• Introducing new projects to business stakeholders (actuaries).

The mission at BPCE-SI takes place within the Regulatory Reporting Management team of the Finance, Risk & Financial Security division. Following a project developed by an external provider; Sopra Consulting, based on artificial intelligence and enabling the preventive detection of anomalies in customer operations.

Missions

• Standardized exchange zones: Populating specifications and business rules for the fields that make up the data bank, based on analyses and UAT.

• Data extraction.

• Data analysis (cleaning, formatting data, etc.).

• Developing scripts and correcting code errors.

• Environment evolution (Local / UAT / homologation).

• Evolution of the anomaly detection engine scopes.

• Deployment.

• Data exposure (visualization, Dashboard).

Within SAS Institute, I was part of the Consulting entity, and also integrated into the team of Groupama Assurance, one of SAS's clients, as part of the anti-money laundering project. As part of this mission, it was necessary to address the client's operational issues in the context of deploying the AML solution. As well as developing a machine learning model to detect high-risk alerts for money laundering.

Missions

• Challenge the AML solution by comparing the results of several machine learning algorithms (Random Forest and Gradient Boosting methods) and testing performance (SAS VIYA).

• Mapping and analysis of the log from a SAS code used to process the customer database of Groupama Assurance.

• Production of a detailed summary of inputs and outputs with mapping rules for each variable.

• Tracking operational alerts of the AML solution at the client's site.

• Construct a flowchart of the AML code algorithm.