Chris
Odhiambo DevOps & Machine Learning Engineer

Designed and implemented a complete CI/CD pipeline for an AliExpress-style e-commerce platform using Jenkins, Docker, and Kubernetes on AWS EKS. Automated testing, deployment, and scaling reduced deployment time by 75%.

Migrated a Vagrant-hosted local application to AWS by lifting and shifting core components. Configured two basic security groups for frontend and backend, launched EC2 instances accordingly, and successfully deployed the application to verify environment parity and accessibility in the cloud.

Migrated to AWS by performing a lift-and-shift: launched EC2 instances in separate security groups for frontend and backend. Finally re-architected the entire infrastructure manual This setup ensured scalability, security, and clear separation of concerns between public-facing and internal components.

Integrated Jenkins as the CI engine for the application. Configured freestyle jobs that were automatically triggered on code commits to GitHub. The pipeline included steps to pull source code, run unit tests, build the Java application using Maven, and publish the artifact to a Nexus repository. This integration ensured build consistency, automated feedback, and early error detection, aligning with CI/CD best practices.

Developed an AI-powered resume optimizer that scans a candidate’s CV and compares it against a target job description to identify keyword gaps, tone issues, and formatting recommendations. The application was built using Streamlit and leverages basic NLP techniques to match resume content with hiring trends.

Designed and deployed a multi-AZ AWS infrastructure using Ansible with dynamic configuration via boto3. The setup included public and private subnets, a NAT gateway, and a secure jump server for accessing private instances. Infrastructure provisioning was automated using Ansible playbooks

Developed a machine learning predictor model on Kaggle The project involved, Multiple models were trained and evaluated, including logistic regression, decision trees, and random forests, with hyperparameter tuning applied for performance optimization. The final model achieved a high accuracy and was validated using cross-validation and confusion matrix metrics.

Provisioned and orchestrated a virtualized multi-VM infrastructure using Vagrant and VirtualBox. The environment simulates a real-world deployment scenario with app servers, a load balancer, and monitoring components.