Platform Engineer

Perry Weather•Dallas, TX

49d•Onsite

About The Position

Perry Weather is the weather safety standard for organizations that operate outdoors. We combine on-site weather hardware with a software platform that automates alerts, siren activations, and safety decisions for more than 3,000 organizations — from the PGA of America, the NFL, and MLB to Turner Construction, thousands of school districts, cities, and golf courses. We've grown 80%+ annually for five consecutive years, and are continuing to grow. We're headquartered in Dallas at The Centrum, in the Oak Lawn neighborhood, where our teams work together every day to help organizations make faster and safer decisions when weather puts people, assets, and operations at risk. We're looking for a Platform Engineer to help build and run the infrastructure the rest of engineering depends on. You'll partner closely with our Lead Platform Engineer, designing and delivering the work together, and own real surface area from the start: infrastructure as code, our Kubernetes environments, delivery pipelines, observability, and the guardrails that keep all of it safe to change. Engineering is your customer. Success in this role looks like fast service setup, uneventful deploys, and engineers who can diagnose a failing workload without asking for help. Engineering here also works with AI coding agents daily, which adds to the platform's job: checks that catch generated infrastructure errors before they apply, scoped credentials and audit trails for automated actors, and cost reporting that accounts for agent usage. You'll use agents in your own work and build these controls for everyone else.

Requirements

4+ years in platform, infrastructure, DevOps, or backend engineering, with real ownership of production infrastructure
Terraform you've written and maintained. You've reviewed a plan and caught something you didn't want applied, and you know how state drift happens
Kubernetes in production. You can take a failing workload, trace it through configuration, networking, and resource limits, and explain what went wrong to the engineer who owns the service
CI/CD you've built and debugged, ideally GitHub Actions. You've fixed a pipeline engineers had stopped trusting, and you can separate a failing test from failing infrastructure
Working depth in a major cloud, including managed databases, networking, identity, and secrets. We're primarily on Azure, and AWS or GCP experience transfers
Scripting in Python, Go, or Bash good enough to automate a manual process end to end and leave it maintainable for someone else
Observability work you've done yourself: instrumentation you added, a dashboard other people used, and an alert you tuned because it fired too often
On-call experience for production systems, including at least one incident you drove to resolution and wrote up afterward
Product instincts about internal tooling. You've built something for other engineers, watched them use it, and changed it based on what you saw
Fluency with AI-native engineering tools and agentic workflows (e.g., terminal-native coding agents, LLM-assisted code refactoring and generation) to multiply technical output and speed up development cycles
Least-privilege credential design for automated systems: CI service accounts, scoped tokens, short-lived credentials, audit trails. Agents are the newest consumers of that work, and the principles carry over
A view on how automated changes should be reviewed. More code and configuration now arrive generated, which puts weight on the checks that run before a merge or an apply, and you should have opinions about what those checks need to cover

Nice To Haves

Experience with policy-as-code
Cloud cost management or FinOps practices
API gateway operations
DNS and CDN configuration
Helm chart authoring
Managed time-series or high-volume databases
Load and performance testing
Running AI or ML workloads on shared infrastructure
Exposure to fleets of connected hardware

Responsibilities

Write and maintain the Terraform that defines our cloud footprint, review changes for blast radius, and keep our modules usable by engineers outside the platform team.
Run our Kubernetes clusters and the Helm-deployed services on them, including resource tuning, scaling behavior, and the failure modes that only show up under load.
Build and maintain CI/CD in GitHub Actions so that builds, tests, and deploys are fast and dependable enough that engineers rely on them.
Improve the instrumentation, dashboards, and alerting that tell us how the platform is behaving, and make alerts specific enough that people act on them.
Track where our cloud spend goes, find the waste, and add controls that catch expensive changes before they ship.
Implement secrets management, least-privilege access, dependency and image scanning, and policy checks that catch mistakes during review.
Give AI coding agents and automation scoped identities and audit trails, add the checks that catch generated infrastructure mistakes before they apply, and track what their usage costs alongside the rest of our spend.
Share the on-call rotation, respond when the platform degrades, and make the change that prevents a repeat.