BRIDGE

• IAM/IGA
• Python
• MongoDB
• Event Sourcing
• CQRS
• LLM/RAG

Designed during my cybersecurity engineering internship at MTN Cameroon and formalized in my Master's thesis (Python, MongoDB, JWT, SFTP). Hybrid deterministic/probabilistic reconciliation, complemented by LLM/RAG assistance on residual cases, reaching 99.5% precision and 99.9% recall on a 5,000-identity annotated sample, benchmarked against Splink. Business Rules Engine catching 100+ non-conformities. Differential auditing via event sourcing with SHA-256 chaining (CQRS pattern). Integrated into an operational application deployed internally at MTN, whose name and deployment details remain confidential. The BRIDGE architecture itself is detailed in my Master's thesis (defended in 2026).