This is a constructive, curated selection of a few enterprise-grade cloud solutions, automation systems, and security initiatives I have built — many of which have been deployed and used at scale.
The repository-available projects below provide visual insight into much of my infrastructure work. Click any project to explore the source, READMEs, and IaC templates on GitHub.
End-to-end, event-driven data pipeline on AWS. Raw CSV uploads to S3 trigger Lambda for validation, which kicks off Glue ETL jobs that transform and enrich data into an Athena-optimized store. QuickSight then renders modern, query-ready business intelligence dashboards. The repo ships with both CloudFormation and Terraform IaC templates for one-command deployment.
Demonstrates the deployment of production-grade Kubernetes (EKS) clusters on AWS using fully modular Terraform. Clusters are provisioned inside a secure VPC, with managed node groups, IAM roles, and networking codified as reusable IaC modules — delivering consistent, scalable, and highly available environments suitable for enterprise workloads.
Reusable Terraform modules for repeatable deploys
Hardened VPC networking and IAM least-privilege
Managed EKS node groups with autoscaling
Built for scalability, availability, and partition tolerance
Architecture Flow
Terraform → VPC + Subnets → EKS Control Plane → Managed Node Groups → Workloads
Full build-and-deploy of a sophisticated GenAI Recommendation Agent using the Strands framework and AWS Bedrock AgentCore. The agent searches the web for relevant content and uses AgentCore Memory to persistently remember user preferences (dietary needs, favorite cuisines, etc.) across sessions. Includes least-privilege IAM via Terraform and notebooks covering AgentCore Identity, Policy, and Gateway concepts.
Persistent cross-session memory via AgentCore Memory
Serverless production deployment with AgentCore Runtime
Least-privilege IAM defined in Terraform
Hook-based context injection and memory writes
Architecture Flow
User → AgentCore Runtime → Strands Agent → Bedrock LLM + Memory + Web Tools
End-to-end CI/CD pipeline of an automated serverless web application powering a live production site (Selah Studios). The project encapsulates the technical details of the web components, build tooling, and infrastructure — including automated build, test, and deploy stages, infrastructure-as-code provisioning with Terraform, and a modern React + TypeScript + Vite frontend.
Serverless infrastructure that automates request submissions and manager approvals. Team members submit requests through API Gateway, which trigger Lambda functions and a Step Functions workflow that notifies managers. Approvals/denials update request state in DynamoDB, with status tracked in real-time through a dynamic queue — a secure, scalable approval microservice with no servers to manage.
Step Functions orchestrate the approval workflow
Real-time status tracking via DynamoDB
Secure API Gateway endpoints with IAM controls
Fully event-driven and horizontally scalable
Architecture Flow
API Gateway → Lambda → Step Functions → SNS Notify → DynamoDB Status
Built at Google, this ML-driven monitoring system leverages GCP tooling and ITSM workflows to detect, triage, and auto-remediate infrastructure inconsistencies in real-time — reducing mean time to resolution and keeping critical services healthy at scale.
GCPMachine LearningITSMPython
Google
CI/CD Infrastructure Optimization
Engineered CI/CD pipeline enhancements at Google using SQL and Python to instrument OKR tracking and automated security checks. The work hardened release safety and observability, contributing to a 40% increase in system uptime across managed services.
CI/CDPythonSQLSRE
Ernst & Young
Cybersecurity Compliance Platform
Led delivery of a cybersecurity compliance platform at Ernst & Young — developing 60+ business use cases and remediating 15+ critical issues across network and data-privacy controls. The program drove a 25% increase in PCI and data-privacy compliance across audited environments.