Built a fully serverless cloud cost optimization platform that applies Scikit-learn Isolation Forest ML models to detect AWS spending anomalies in real time. Automated EBS/EC2 cleanup triggered by EventBridge, with SNS alerting and DynamoDB audit trails — cutting projected cloud waste by 30%.
Built an end-to-end automated CI/CD pipeline using GitHub Actions. Orchestrated containerized deployment flows using Docker and Kubernetes clusters on AWS EC2 for continuous, zero-downtime delivery.
Built a highly scalable serverless e-commerce platform using AWS Lambda and API Gateway. Integrated DynamoDB for product/order data management and SNS for real-time order notifications.
Containerized a Python FastAPI microservice using Docker and deployed it onto a Kubernetes cluster hosted on AWS EC2, demonstrating production-grade high availability and automated horizontal scaling.
Configured EventBridge rules to listen to EC2 state-change events and trigger Lambda functions that push real-time alerts via SNS. Reduced incident response time from 10+ minutes to under 30 seconds.
Wrote Python ETL scripts to parse and migrate data from S3 JSON files into DynamoDB tables automatically. Configured Global Tables across ap-south-1 and us-east-1 for <100ms latency worldwide.
Cleaned and preprocessed raw sales data using Python and Pandas, performed statistical analysis and EDA to identify key patterns and trends, and built a predictive model using Scikit-learn to forecast future sales.
Built a data pipeline moving raw inventory data from S3 to HDFS for storage, processed and queried using Hive to analyze stock and demand trends, then loaded results back to S3 and visualized inventory insights in Power BI.