MILHAN ZAHID

• Architected and deployed a full-cycle AI-powered medical billing system using Python and
LangChain, engineering intelligent automation pipelines that reduced manual intervention
across critical revenue cycle touchpoints.
• Implemented an automated denial management and claim validation engine using RAG-based
validation logic, driving sub-5% denial rates and 90%+ collection efficiency against an industry
average of 60-80%.

• Trained ML models with accuracy gains of +18% over baselines using Scikit-learn and Vertex
AI, leveraging AWS (S3, Lambda, Redshift) and GCP (BigQuery, Vertex AI) to achieve 99.9%
uptime reducing infrastructure costs by 25%.
• Architected and implemented automated monitoring and model retraining pipelines using
MLflow and Evidently AI to detect data drift and concept drift, ensuring ML models remain
reliable and effective over time.

• Engineered end-to-end automation pipelines integrating GCP services (BigQuery, Cloud
Storage) with Python-based orchestration, reducing manual data processing time by 40%
across business workflows.
• Automated monitoring systems using Prometheus to track pipeline health and data quality,
flagging anomalies proactively and maintaining 99%+ data accuracy across production
environments.

• Designed and deployed automated data ingestion pipelines using Python and BigQuery,
processing 10M+ records daily with 99%+ pipeline reliability across production environments.
• Established scalable analytics workflows from scratch using dbt, defining data requirements
and owning delivery end-to-end.