Senior Platform Engineer (f/m/x)

Your role with us As Senior Platform Engineer , you own the platform that Workist's AI agents run on: two clouds, the Kubernetes clusters for production, staging and ML workloads, and the path from merge request to production for every service we ship. You decide how this platform evolves, you keep it secure, observable and cost-efficient, and you extend it as the product grows. Most recently that meant an LLM gateway with region failover in front of our Azure OpenAI deployments, and performing GPU capacity planning for our own models. You set your own roadmap. Infrastructure at Workist is run as quarterly themes that you propose, make the case for and deliver, and roughly a third of your time goes to whatever the week brings: an incident, a pentest finding, a developer whose deployment is stuck and needs a second pair of eyes. You report to our CTO and are the voice of the platform in engineering decisions. How we build and run things Everything runs on Kubernetes (Azure AKS), deployed with Helm and GitOps/Flux. All infrastructure is Terraform, applied through CI. CI/CD for every service on GitLab Managed services wherever they keep life simple: Postgres, OpenSearch, Redis, blob storage. We self-host only where it clearly pays off. Two clouds: Azure as our primary cloud, AWS for search and mail ingestion. Python everywhere: read and fix application code when that is where the fix belongs. Security is routine: automated scanning in every pipeline, regular external pentests, quarterly backup and disaster-recovery tests. Your responsibilities Own the platform Own the architecture of our cloud environments across Azure and AWS: how subscriptions, networks, identities and environments are structured and stay isolated Design and deliver the platform capabilities the product needs next, from GPU capacity and an LLM gateway to new environments Set the standard for how we run Kubernetes and our managed data services: capacity planned ahead of growth, changes and upgrades rolled out safely and routinely, costs kept in check Keep it quiet Own the security posture end to end: least-privilege access, network isolation, findings from scanners and pentests, and a patch cadence that runs itself. Be a technical counterpart for audits and security questionnaires Own observability and alerting so that every alert is worth acting on, and lead the response to platform incidents including the follow-up fixes Run quarterly backup/restore and disaster-recovery tests, and close what they reveal Make the team faster Own the path from merge request to production: fast builds, reliable deployments, environments and access on demand Build the infrastructure behind our AI models: the LLM gateway, model deployments with region failover, GPU capacity, and whatever the next model needs What this looked like recently : a Redis instance that silently dropped half of its new TCP connections and took the Celery workers down with it, region failover and load balancing for our Azure