Discord

Data Scientist - Client Platform

Analytics-focused data scientist on Discord's Client Platform team, improving mobile and desktop reliability and app performance through metrics, experimentation, and dashboards.

The role

This role sits within the Client Platform Data Science group and supports the engineers and product teams who build and run Discord's apps. The focus is on keeping the client experience dependable and improving it through analysis that spans exploration, dashboards, and controlled tests. The person will also help shape the instrumentation and self-serve tools that other teams rely on.

What you would do

  • Work with teams across the company on analyses, reporting, and experiment readouts
  • Set key and diagnostic metrics for user experience and build dashboards that stay current and useful
  • Construct bespoke datasets to track new features and workflows
  • Present findings to both technical and general audiences and refine them using their input
  • Guide teams on sound experiment design, analysis, and interpretation
  • Partner with data and engineering to plan durable logging and self-service analytics tools

What they asked for

  • At least two years of hands-on analytics work resolving unclear business questions into concrete recommendations
  • Experience designing, analyzing, and reading A/B tests at large scale
  • Solid SQL and skill building fast, well-structured dashboards in tools like Tableau or Looker
  • Ability to explain complicated results in plain terms to varied audiences
  • Inquisitive mindset, teamwork orientation, and persistence on difficult problems

Nice to have

  • Enthusiasm for Discord or online community products
  • Background in measuring performance or reliability for online services
  • Familiarity with social or subscription metrics such as network effects, lifetime value, or conversion funnels
  • Experience with production ETL pipelines or BigQuery
  • Use of AI tools in analytical work
Selected employer postings. AI-assisted summaries may contain errors. This is not a market-wide sample. Independent of the companies listed.