Hamza Faidi

Designed & led Mastercard virtual-card renewal via GraphQL banking APIs, a scheduled worker on Azure Queue Storage orchestrated with Azure Logic Apps. Continuously monitored rollout; renewed 100+ cards before expiry, protecting interchange revenue on card transactions (business-critical). Built a multi-layer, agentic fraud detection system for Mastercard transactions, combining mult layer LLM analysis to flag suspicious activity in near real time. Implemented backend in .NET with a fraud-review UI in Angular. The system achieved ~79% fraud detection accuracy in production, halved manual reviews, and recovered €3,000+ in fraudulent charges. Drove provider-agnostic payments refactor to integrate Adyen cleanly. Applied Observer/Factory/Decorator patterns, improving maintainability, ~30% faster feature delivery, and ~20% faster onboarding. worked on a strategic project to migrate legacy code to a Domain-Driven Design (DDD) architecture, which included a complex database schema migration. Authored cautious SQL scripts to avoid resource locking and minimize performance impact in production. Prepared and monitored migration jobs with detailed logging to detect failures early, and completed migration with zero downtime and <2% CPU overhead during peak loads

Developed an automated credit card suspension feature for Expensya’s finance platform, replacing a manual support process. This reduced fraud risk windows and improved card availability for 1,000+ corporate users (administrators can set rules to auto-suspend cards if receipts are missing for too long). Designed the end-to-end workflow: I built a policy management UI in Angular, created secure ASP.NET Core APIs for execution, and used Azure Logic Apps plus Queue Storage for orchestration and retries. This ensured reliable, scalable processing across multiple companies. Enhanced security by integrating Azure Key Vault for managing secrets (API keys, tokens) in the card lifecycle process, reducing exposure of sensitive data.