The Role
This role sits at the intersection of DevOps and ML infrastructure – embedded within the ML team, not adjacent to it. The immediate focus is hands-on and concrete: a GitHub migration with a full CI/CD refactor, Kubernetes scaling work under real ML workloads, LLM infrastructure management, FinOps and cost optimization, and the groundwork for multi-cloud expansion. You’re not supporting the platform from a distance – you’re a working part of the team that runs it.
About the Product
The platform processes live sports broadcasts from 250+ leagues and broadcast partners – including the NBA, NHL, ESPN, and Bundesliga – and automatically generates short-form highlight videos in real time. It runs on Computer Vision and NLP pipelines at cloud scale, with no tolerance for downtime. Infrastructure decisions here have direct product impact: there’s no abstraction layer between what you build and what the platform delivers.
The Stack: Infrastructure runs on Kubernetes, provisioned via Terraform, with CI/CD through GitHub Actions. Observability is covered by Datadog and/or Grafana. The ML workloads span Computer Vision, NLP, and Data Science pipelines – you won’t be building the models, but you’ll own the infrastructure that keeps them running reliably at scale.
What You’ll Be Doing
- Lead the GitHub migration and CI/CD refactor – redesign release pipelines, deployment gates, and source control workflows end-to-end
- Own Kubernetes in production: cluster health, workload scheduling, scaling, and reliability under ML workload conditions
- Manage LLM infrastructure: model serving pipelines, resource allocation, and serving reliability
- Drive FinOps practices – identify cost inefficiencies, implement optimization across the platform
- Lay infrastructure groundwork for multi-cloud expansion
- Write and maintain Terraform configurations across the platform
- Close collaboration with the DevOps Team, Data Science, ML, and Algorithm teams
What We Expect
Must-have
- 5+ years of commercial Kubernetes experience in production environments
- Demonstrated ability to design infrastructure solutions, not just operate them
- 3+ years of commercial Terraform experience
- Hands-on GitHub Actions experience (or equivalent CI/CD tooling at production scale)
- Working knowledge of Datadog, Grafana, or comparable observability platforms
Nice to have
- Kubernetes at high scale – high RPS, large cluster counts, multi-tenant workloads
- Background in ML infrastructure: experiment tracking, model registry, serving pipelines
- Azure Cloud experience in production
- Experience with Coding Agents, MCP servers, or AI-assisted DevOps tooling – commercial or self-driven
Why This Role Is Worth Your Time
The product is technically genuine: real-time AI on live sports data, serving some of the biggest names in the industry. Infrastructure work here directly shapes what the platform can do – the feedback loop between decisions and outcomes is short. The team operates in a hybrid model from Warsaw. If you’re someone who engages with AI tooling not just as a user but as a builder, that orientation is actively valued on this team.
