Data scientist and business operations leader with an M.S. in Applied Data Science and 10+ years of retail leadership experience. I build predictive models, ETL pipelines, and interactive dashboards using R, Python, and various data-visualization tools. My work spans retail optimization to research on economic resilience, translating complex analyses into actionable insights that drive strategy.
In additon to highllighting specific projects, this portfolio website is also a live demonstration of my skills in DevOps, cloud automation, and reproducible publishing workflows.
It is generated using Hugo, a fast static site generator written in Go, and is designed to be fully version-controlled, reproducible, and automatically deployed via a cloud-native CI/CD pipeline.
Project at a Glance
This site pipeline demonstrates:
Infrastructure-as-code and CI/CD using AWS CodeBuild, IAM, and GitHub
Fully automated deployment of a static site to a CDN-backed global endpoint
Content-driven engineering using Hugo and modular markdown structures
Security-aware design with scoped IAM roles and reproducible builds
Pipeline Overview
1. Local Development
Content is authored in markdown using Hugo’s content/ structure and organized by theme-defined sections (e.g., about, portfolio, projects)
Theme customization is handled via the creative-portfolio-mb Hugo theme with parameters set in config.toml
Pages are previewed locally using hugo server for rapid iteration
2. Version Control & GitHub Integration
The entire site (content, configuration, theme, assets) is tracked in Git and hosted in a GitHub repository
Commits to the main branch represent production-ready changes
3. Continuous Deployment via AWS CodeBuild
A GitHub webhook triggers an AWS CodeBuild project whenever a push is made to main
The build process is defined in a buildspec.yml file, which includes:
Installing Hugo
Building the static site with hugo
Deploying the site to an Amazon S3 bucket configured for static website hosting
Invalidating the CloudFront cache to ensure instant global propagation of updates
4. Secure & Automated Delivery
Deployment credentials are scoped via IAM to permit only S3 and CloudFront actions
AWS CLI is optionally configured locally to allow manual hugo deploy overrides
Assets are delivered over HTTPS via CloudFront with long-term caching policies defined in config.toml