Pavan Kumar Adapala

- Designed production-ready MLOps pipelines with MLflow and DVC, enabling reproducible training, and model versioning across multiple ML projects. - Automated the ML delivery workflow by integrating GitHub Actions, Terraform, and API endpoints, reducing manual deployment effort and improving time-to-production for ML applications by 40%. - Engineered Kubernetes-based deployment environments for containerized ML inference services, enabling reproducible deployments and scalable model serving.

Designed industrial machine learning solution for HVAC energy load forecasting using IIoT time-series data, supporting Predictive Energy Management within automotive paint shop operations. Developed and evaluated ML forecasting models across different data aggregation levels, analysing model performance to identify suitable forecasting approaches. Improved operational energy planning and load-shifting decisions under varying weather conditions by validating forecasting models against real industrial data.

- Optimized nightly and integration-test Jenkins CI/CD pipelines for distributed ADAS software development, improving build reliability and deployment efficiency. - Developed Python-based automation functions integrating REST APIs and web automation into Jenkins CI/CT/CD workflows, automating test reporting and reducing debugging time by 30%. - Standardized CI/CD observability across multiple ADAS projects through centralized Grafana dashboards, enabling engineering teams to identify release bottlenecks.

Designed and maintained standardized AWS cloud platforms using Terraform, reducing infrastructure provisioning time by 35% and enabling development teams to deploy applications reliably. Engineered enterprise CI/CD pipelines for microservices deployment using Jenkins, adopting a shift-left delivery approach by integrating automated code quality analysis and Docker image builds. Containerized applications using multi-stage Docker builds and orchestrated deployments with Kubernetes.

Contributed to an enterprise AWS cloud migration initiative by designing scalable 3-tier cloud architecture with Auto Scaling and High Availability, reducing infrastructure costs by 25%. Established GitHub repository governance through branch protection policies and mandatory CI checks, improving code quality and reducing integration issues. Engineered Jenkins-based CI/CD pipelines using a master-agent architecture to automate software deployment and distribute build workloads, reducing build times by 20%.