A comprehensive demonstration of modern DevOps practices, this project showcases my expertise in cloud-native application deployment. I engineered a fully automated 3-stage CI/CD pipeline leveraging AWS CodePipeline, CodeBuild, CodeDeploy, CodeArtifact, IAM, and CloudFormation to enable zero-touch deployments directly from GitHub to EC2 instances.
The solution features secure package management and infrastructure-as-code principles, resulting in deployment times under 5 minutes per code push. This project highlights my ability to architect end-to-end DevOps solutions that balance security, efficiency, and reliability.
At the intersection of blockchain security and machine learning, I developed an advanced phishing detection system for Ethereum transactions under the guidance of Dr. Vishnu Srinivasa. This system processes a massive dataset containing 13.5M+ transactions and 2.9M nodes, using dimensionality reduction via PCA to enhance model performance.
I implemented and compared multiple ML models (SVM, Random Forest, XGBoost, and MLP), achieving 96.78% accuracy and 88.49% F1-score with the MLP model. The system is optimized for real-time security applications, demonstrating my ability to apply AI techniques to solve complex cybersecurity challenges in blockchain environments.