Infrastructure-as-code platform enablement

Infrastructure-as-code work across Understory's core Terraform repositories, provisioning the DynamoDB tables and access patterns, IAM permissions, and Glue/Athena cataloging infrastructure that the rest of the company's services ran on top of.

Concretely: bringing existing DynamoDB tables under Terraform management, adding indexes so services could query by the fields they actually needed instead of scanning, and setting up Athena catalogs backed by Glue crawlers so newly landed data became queryable without a manual cataloging step. When the lead-sourcing platform needed its own infrastructure, I provisioned it directly in Terraform rather than clicking through the AWS console, and later helped shift its monitoring app onto the same kind of infrastructure too.

This is the layer of the job that's invisible when it's working and very visible when it isn't, and coordinating changes to it across a team meant a steady rhythm of small, reviewed pull requests rather than big infrastructure rewrites landing all at once.