At MTN Fintech, I built a Python reconciliation engine that analyzes breaks, tags root causes, and categorizes exceptions, achieving over 95% auto-matching accuracy on production runs. I also developed parsers for MT940, ISO 20022, CSV, and XLSX financial statements.
At Zulaiy Technologies, I built automated ETL pipelines processing over 500K records daily and a Gemini-powered OCR pipeline for extracting e-invoicing data. I also prototyped an AI decision layer for an e-governance platform.
At Noji Guardian Protocol, I designed and maintained cloud infrastructure, reducing monthly costs from over $600 to about $40. My work spans data engineering, cloud infrastructure, and AI-assisted data systems.
