Sr Cloud Platform Engineer

Passively Looking

DevOps / SRE Stamford, Connecticut, United States

Profile
Contacts

Summary

Most recently, built an automated pipeline in AWS Code-Pipeline that used a split Terragrunt+Terraform infrastructure. Aside from the usual connecting of the pieces between existing software and services of AWS, Github, Terraform, and Terragrunt, I wrote a custom python script that matched Terragrunt modules to their infrastructure paths, that allowed for different behavior of the main-org account resources (which was build, approve, then deploy -- these resources had no notion of staging/production, and included things like pipelines that built other pipelines) and the resource-account resources, which were the accounts and environments where the resources deployed lived -- in staging and production. The pipeline infrastructure was also designed for developer interaction -- abstracting away most of the terraform knowledge that would be needed into YAML files. This approach could also make it possible to automate further -- instead of generating TF for the pipelines, it would be possible to generate the YAML.

Prior to that, the thing I was most proud of creating was a lambda deployment pipeline that utilized s3 as a back-end resource, kind of a poor man's artifactory.

Expectations

Development/devops roles using scripting/automation, where I can bring my passion for quality and DevOps theory to a dynamic team oriented environment. Excited about working with open-source technology and contributing to the community. Very comfortable working independently -- can innovate solutions -- including envisioning, designing, and implementing POC, and then creating pipelines to deploy, -- but also greatly appreciate more eyes and minds for those times when I get too myopically invested in an approach, or when it isn't cost-effective for me to consume time figuring out how a thing works.

Culture is important. Particularly the way that neurodiverse people are treated. I work best in environments where context switching is minimized as much as possible.