MILHAN ZAHID

Architected and deployed a full-cycle AI-powered medical billing system using Python and LangChain, reducing manual intervention across revenue cycle touchpoints. Built a RAG-based denial management and claim validation engine achieving sub-5% denial rates and 90%+ collection efficiency (vs. 60–80% industry average).

Engineered end-to-end automation pipelines integrating GCP (BigQuery, Cloud Storage) with Python orchestration, reducing manual data processing time by 40% across business workflows. Implemented Prometheus-based monitoring to proactively flag anomalies and maintain 99%+ data accuracy in production.

Designed and deployed automated data ingestion pipelines with Python and BigQuery, processing 10M+ records daily with 99%+ pipeline reliability. Built scalable analytics workflows from scratch using dbt and owned end-to-end delivery of data requirements.

Trained ML models to improve accuracy by +18% over baselines using Scikit-learn and Vertex AI. Implemented automated monitoring and model retraining pipelines with MLflow and Evidently AI to detect data and concept drift, achieving 99.9% uptime and reducing infrastructure costs by 25%.

Optimized monthly financial reconciliation by building automated BigQuery data pipelines, cutting processing time by 30%. Created KPI dashboards using BigQuery and Data Studio to deliver real-time performance visibility with zero manual refresh cycles.

Earned a bachelor’s in Data Science from Islamia University in Bahawalpur. Coursework supported skills used in later AI and automation work.