At Impetus technologies, I designed and automated end-to-end ML pipelines using Airflow (GCP Composer), PySpark, and Python to handle model retraining and deployment. I also optimized an XGBoost model for better execution efficiency and standardized ML metrics so evaluation stayed consistent across use cases.
I built a centralized hyperparameter tuning framework to make optimization scalable and reusable across the ML platform. By parallelizing data transformations and optimizing pipeline scheduling, I improved model execution speed by 80%.
I’ve also developed CI/CD pipelines with Jenkins and GitHub Actions to streamline testing, deployment, and release workflows, and I’ve delivered infrastructure automation—18+ ETL templates, Terraform-driven deployments, and a high-performance, secure PySpark solution on AWS.
