Azure AI Platform Engineer III/Senior
EQT Corporation, Remote
The role
At EQT, you won't just do a job, you'll help shape our future. Through Transparent Leadership, you'll work alongside leaders who value collaboration and ensure every voice is heard. Through Connected Impact, you'll see how your work influences critical business decisions and drives value across our operations. Through Supported Growth, you'll have opportunities to expand your skills through cross-functional collaboration, development opportunities, and career advancement.
Join our Qrew!
EQT runs AI agents in production on Microsoft Foundry, working inside Salesforce and the enterprise systems our people use every day. Demand is compounding as domain experts across the business bring new agent ideas forward, and the platform underneath has to scale with it.
We need an Azure AI Platform Engineer III/Senior to own that platform: secure, observable, affordable, and easy to build on.
You own the foundation every agent runs on. Our AI Engineers build agents and the tools they use; you build and operate the layer beneath them: the Azure environment, the AI gateway, the tool registry, and the observability, evaluation, and guardrail systems that keep the fleet healthy.
This is a senior, hands-on engineering role. You write the infrastructure as code, own the AI platform in production, and partner with Infrastructure to align to standards and deploy. You won't spend most of your time building agents, but you'll understand them well enough to make the platform fast and safe to build on.
What you will do
- Own our Azure infrastructure as code: author, review, and deploy modular templates across dev and production.
- Operate Microsoft Foundry environments, model deployments, and our private Azure subscription as production infrastructure, including onboarding, upgrading, and retiring models.
- Own the AI gateway on Azure API Management: routing and transformation policies, per-consumer keys, rate limits, and onboarding new models and consumers.
- Run the centralized tool registry: identity and authentication patterns, versioning, and reuse standards, so a tool built once serves every agent.
- Build fleet-wide observability: logging, compliance log collection, monitoring, alerting, and usage telemetry attributed by agent and user.
- Build and operate the evaluation infrastructure that gates releases and continuously scores agents in production.
- Own identity and access: Entra ID group-based, least-privilege RBAC, managed identities, just-in-time elevation, and drift detection against the code.
- Own cost and performance: