Rajaofera Ari

• Developed a Power BI dashboard comparing the performance of sensor technologies (old vs. new) and monitoring the real-time status of over 500 sensors across 6 subsidiaries. • Enhanced data collection by adding technical performance data through meetings with international subsidiaries, contractors, and vendors. • Centralized of over 10 million data points from various sources into a single database, improving accessibility and analysis. • Applied the Kaplan-Meier statistical method to estimate the percentage of surviving sensors.

• Model training to detect unpaid customer invoices, helping to reduce default losses. • Fine-tuning BERT models to analyze the sentiment of audit reports and assist the analyst's decision to accept customers for credit. • Creation of an interactive Power BI dashboard to help the team process the various mail flows.

• Use of clustering models to identify vehicle anomalies and anticipate possible technical failures.

• Implementation of a Java application so that developers can retrieve their code using REST Web services. • Code modernization: obsolete methods replaced and XML to JSON conversion added.