Principal Investigator · Completed
AI-based Service Request Analysis and Rate Limiting
Led an R&D project (RP #N-IS1-25-0006) for the National Database and Registration Authority (NADRA) on fair and efficient use of national service resources. Built an unsupervised anomaly detection model over engineered behavioural features from roughly 671,000 service requests, and designed a behaviour-driven rate-limiting framework that maps continuous anomaly scores to graduated risk controls. The retrainable framework, trained models, and validation pipeline were delivered to NADRA in November 2025.