Muhammad Ali

Built AWS ETL pipelines for time series building sensor data feeding production ML systems. Processed parallel streams with PySpark, improving throughput by 20%. Orchestrated workflows in Airflow with retries, alerts, and dependency handling. Deployed Docker/Kubernetes services on AWS and monitored data quality, drift, and system health. Automated ML CI/CD with GitHub Actions and managed Linux EC2 servers, logs, cron jobs, and documentation.

• Analyzed network performance and outage reports to identify root causes and improve reliability. • Tracked operational costs and visualized metrics using dashboards and reports. • Delivered actionable, data-driven insights to technical teams to reduce downtime and optimize operations.

1. Designed and implemented an ETL pipeline for high-frequency PV and transformer data. 2. Managed and optimized data warehouses, enabling real-time analytics and reporting. 3. Built and deployed forecasting models for solar energy production. 4. Automated data ingestion and transformation workflows to improve pipeline efficiency. 5. Collaborated with cross-functional team to ensure data quality, consistency, and scalability.

Thesis Title: Development of a Scalable Machine Learning Pipeline on AWS for Optimizing Building Heating Systems using Reinforcement Learning and Transfer Learning Skills Developed: Machine Learning & AI, AI Agents, RAG, AWS, Cloud Computing, MLops, Databases, ETL Pipelines