Projects

AI-Driven MLOps Pipeline with End-to-End Automation Introduction Machine Learning models are powerful, but deploying them at scale while ensuring automation, reproducibility, and performance monitoring is a challenge. This project aims to build an AI-Driven MLOps Pipeline that seamlessly integrates Kubernetes, Terraform, CI/CD, and cloud services to automate the entire ML lifecycle. Project Goals Automate ML workflows using Kubeflow. Deploy ML models on Kubernetes for scalability. Use Terraform for infrastructure automation. Implement CI/CD to enable continuous deployment of ML models. Monitor models using Prometheus, Grafana, and MLflow. Technology Stack Category Tools & Services Cloud Platform AWS (EKS, S3, Lambda, SageMaker) Infrastructure as Code Terraform MLOps Orchestration Kubeflow, MLflow Containerization & Deployment Docker, Kubernetes CI/CD GitHub Actions, ArgoCD Monitoring Prometheus, Grafana Programming Python ...

Health-Informatics - Research Project at UIC Analyzed the Illinois in/out patient data acquired by COMPdata informatics & statistically matched the patient records with household population records from the RTI US Synthetic Household population database. Developed ML predictive classification model with random forest giving accuracy of 72% and recall of 78% to predict patients diagnosed with Lung cancer, Breast Cancer, Opioids-overdose etc based on ICD-10 codes.