Nate Swenson

Senior DevOps Engineer · AI Agents · Platform

↓ Download PDF

Summary

Sixteen years building the platforms other engineers ship on. Right now that means AI agents on the Anthropic Claude platform that triage incidents, debug failures, and make the CI/CD calls that used to wait on a human. Before that: AWS at scale, Terraform-driven infrastructure, and the pipelines two major product launches ran on. I like leaving teams faster than I found them.

At a glance

Years
16 years
Cloud and platform engineering
Focus now
AI agents
Incident triage, debugging, CI/CD
Cloud
AWS at scale
EKS, ECS, Lambda, S3, DynamoDB
Reach
100+ engineers
Standards adopted org-wide

Skills

AI & Agents

Anthropic Claude SDK, Model Context Protocol (MCP), LLMOps, agent design and evaluation

Cloud

Amazon Web Services (AWS): EC2, EKS, ECS, Lambda, S3, DynamoDB. Kubernetes, Docker, OpenShift

Infrastructure as Code

Terraform / OpenTofu, Ansible & Ansible Tower, CloudFormation

CI/CD

GitHub Actions, GitLab CI, Jenkins, pipelines-as-code, golden pipelines

Reliability & Security

Datadog observability, Site Reliability Engineering (SRE), on-call, SAST/DAST, secret scanning, policy-as-code

Languages

Python, JavaScript, Bash, SQL, Linux

Experience

Senior DevOps Engineer

Nov 2022 – Present

GoodLeap / Remote

  • Co-developing AI agents on the Anthropic Claude platform that automate incident triage, debugging, and CI/CD failure analysis across GitHub. Cut mean time to resolution by 3× in early rollout.
  • Building an agent that auto-tags legacy AWS resources across every account in the org, reconciling infrastructure created by Terraform, SST, CloudFormation, and manual console work into one accurate inventory.
  • Built a repository creation agent that applies company-wide naming, structure, permission, and CI/CD standards to every new repo at creation time, so compliance is enforced up front instead of audited after the fact.
  • Led AWS infrastructure engineering for two major product launches, designing the Terraform modules and GitHub Actions pipelines they shipped on, then set the org-wide CI/CD and IaC standards other teams now build against.
  • Built the Datadog observability dashboards and alerting the on-call rotation runs on today, standardizing how production incidents are detected and escalated.

Lead Associate, DevOps Engineering

Jun 2021 – Nov 2022

Fannie Mae

  • Owned GitLab platform infrastructure end to end at a federally chartered mortgage finance company: Terraform-managed deployment, version upgrades, multi-region high availability, runner autoscaling, and DR/backup automation, all under financial-services security and compliance constraints.
  • Designed and shipped shared CI/CD templates and golden pipelines that engineering teams adopted as their delivery baseline, replacing per-team pipeline drift with one vetted, reusable path.
  • Wired SAST, DAST, secret scanning, and policy-as-code directly into the platform’s pipelines so security and audit requirements shipped with the code instead of after it, then onboarded and mentored the teams adopting it.

Lead Cloud Platform Engineer

Dec 2020 – Jun 2021

Discover Financial Services

  • Architected an automated IAM policy assignment system spanning AWS Lambda, DynamoDB, Jenkins, and Ansible Tower, managing access policies for roughly 300 S3 buckets and giving Discover a self-service path that replaced manual provisioning.
  • Designed and published 20 reusable Ansible templates that standardized provisioning across cloud and on-premise services, adopted by an org of 100+ platform engineers as the vetted baseline to build on.
  • Led the Ansible Tower 3.6 to 3.8 upgrade across 15 consuming teams and delivered the migration without disrupting production workflows.

Lead DevOps Engineer

Sep 2019 – Dec 2020

Optum / Eden Prairie, MN

  • Led DevOps teams supporting government and military health contracts, delivering CI/CD and infrastructure under regulated, compliance-driven constraints.
  • Re-architected Jenkins onto container-based infrastructure, replacing static-VM build agents with elastic on-demand workers. Build throughput doubled and operational downtime dropped by a third.
  • Built zero-downtime automated deployment pipelines for high-availability healthcare workloads, and operated Jenkins, Chef, and GitHub Enterprise enterprise-wide, setting the reliability standards downstream teams depended on.

Senior DevOps Engineer, Community Lead

Jan 2019 – Sep 2019

Optum / Minneapolis, MN

  • Led a distributed team building internal Agile and engineering communities of practice across a 235,000-person organization, scaling adoption of best practices into every product group.
  • Developed and delivered hands-on technical training in TDD, pair programming, and XP, a curriculum that shaped engineering culture across the org.

DevOps Engineer

Jan 2016 – Jan 2019

Optum / Minnesota

  • Built end-to-end test automation across Selenium, Cucumber, and Sauce Labs, replacing manual QA cycles with continuous browser-coverage testing that ran on every build.
  • Shipped zero-downtime deployments on OpenShift and developed pipelines-as-code in Jenkins, giving product teams a version-controlled, reusable release path.

Open source

Seven independently versioned Claude Code skills in one monorepo, each with its own test suite, offline evaluation baseline, and required CI check. Protected promotion flow with auto-merge on green and version-driven release tagging.

A self-hosted Model Context Protocol server exposing 44 grounded tools over a local SQLite store, so an agent has to call a tool before it can state a number. Python and Docker, 85% coverage gate in CI, MIT licensed.

natejswenson.com

Engineering writing

How-to guides on platform engineering and agent tooling, written from real shipped releases.

Education

Bachelor of Science, Industrial Engineering

University of Minnesota Duluth / 2005 to 2010