Member of Technical Staff (Product Data Scientist, Search Quality)
Generate a McCoy IQ challenge in 30 seconds.
See how candidates think and approach the work this role demands, before the phone screen. We'll build a video challenge from this posting, and you can edit or share it before it goes live.
Key details
Job Description
Perplexity is looking for an experienced Product Data Scientist to accelerate the development of advanced search technologies. You will identify robust and sensitive signals from user behavior to help us gather insights from A/B experiment data more efficiently.
Responsibilities
Develop data-driven insights from user behavior to inform our product roadmap and accelerate adoption
Formulate hypotheses and validate them by designing, running, and analyzing A/B tests
Determine appropriate metrics and visualizations for tracking, and implement them in dashboards
Design new pipelines that will help to deliver better ranking quality. From discovering new signals, producing metrics and construct data labeling pipelines with human and LLM feedback
Qualifications
4+ years of experience working as a data analyst or in a related role
Experience working on search-related products, with emphasis on designing online metrics and analyzing A/B experiments
Strong Python skills (expected to write production-grade code)
Proficiency with SQL
Experience with Business Intelligence (BI) tools
Deep knowledge of statistics
Preferred Qualifications
Proficiency with Apache Spark
Experience with Databricks
Experience with development of LLM-as-a-judge systems
Audit details(provenance, verification trail, raw fields)
Core fields
perplexity:a805e14b-061d-469c-9136-b9e6a1855902Provenance
perplexityVerification trail
- confirmed_closed2026-08-01 10:17:41Zvia ashby
evidence
{ "id": "a805e14b-061d-469c-9136-b9e6a1855902", "slug": "perplexity", "signal": "graphql:jobPosting:null", "corroborated_by": "posting_api:absent" }
LLM enrichment
See how we measure for definitions, or our corrections log for known issues. Found something wrong? Flag a correction.