Data Scientist II – Performance Optimization Squad

The Performance Optimization Squad is a newly formed team in the Core Infrastructure Studio with a mission to establish a core competency in performance engineering and address systemic inefficiencies across Spotify's platform. With 3,200+ microservices, 40,000 VMs at peak, and 500,000 K8s pods, even minor fleet-wide efficiency improvements result in substantial cost savings. We've already identified $8M+ in annual savings from our first initiative alone, and we're just getting started.

This squad works horizontally across the entire stack — but none of that optimization happens without the data to see it. We're looking for a Data Analyst who will build the measurement foundation that drives every decision we make.

What You'll Do

Optimization at this scale is only possible when someone can see the problem clearly. You'll build the data foundation from scratch; designing and owning the datasets, pipelines, and metrics that make performance inefficiencies visible across the platform.

Design, build, and maintain datasets and data pipelines that surface resource utilization, cost, and performance signals across Spotify's infrastructure

Define and own metrics for efficiency, latency, and resource utilization;  turning raw infrastructure signals into insights that drive prioritization

Proactively investigate performance data to surface optimization opportunities, not just respond to engineering requests

Build dashboards and analyses that support decision-making across the squad and partnering platform teams

Work with engineers and platform teams to define guardrail metrics, validate findings, and measure the real-world impact of optimization efforts

Translate complex infrastructure data into clear stories for both technical and non-technical audiences

Own the data foundation: There is no inherited data infrastructure here;  you'll design and build it from scratch. What gets measured, and how, is yours to define

See your impact directly: Every insight you surface translates into cost savings. We measure success in dollars saved and efficiency gained;  your work shows up in production

Breadth at scale: Work across the entire Spotify platform;  3,200+ microservices, 40,000 VMs, 500,000 K8s pods. Few companies offer data problems at this scale

Greenfield from day one: Help shape the culture, tooling, and data strategy of a brand new squad with strong executive support

Who You Are

You have experience working with infrastructure, platform, or cloud cost data;  Kubernetes metrics, cost attribution, utilization signals, or observability data feel familiar
Or you're a strong technical analyst with enough grounding in distributed systems and cloud infrastructure to navigate GKE cost data, JVM metrics, and resource utilization signals
You write clean, efficient SQL and Python;  comfortable enough to model data and build lightweight pipelines, not just query existing tables
You're self-directed: at your best when hunting for problems in the data, not waiting to be handed them
Comfortable with ambiguity and able to carve your own path in an early-stage, unstructured environment
Experienced with data visualization tools (Looker or similar) and know how to make a dashboard tell a story, not just display numbers
You communicate clearly and confidently with engineers and non-technical stakeholders alike

Where You'll Be

This role is based in Stockholm or London.

We offer you the flexibility to work where you work best! There will be some in person meetings, but still allows for flexibility to work from home.