YO IT Group
Qualifications
0–2 years in data science/analytics or similar; BS/BA in a quantitative field (or equivalent work).
Strong SQL; Python for analysis; comfort with experiment design and causal thinking.
Communicates crisply with engineers, PMs, and leadership; turns analysis into action.
Nice-to-haves: dbt, dashboarding (Hex/Mode/Looker), marketplace or search/recommendation metrics, LLM/agent evaluation.
What you’ll do In Your First Year You’ll Ship Analyses And Experiments That Move Core Product Metrics—match Quality, Time-to-hire, Candidate Experience, And Revenue. You’ll
Define north-star and feature-level metrics for our ranking, interview analytics, and payouts systems.
Design/run A/B tests and quasi-experiments; turn results into product decisions the same week.
Build source-of-truth dashboards and lightweight data models so teams can self-serve answers.
Instrument events with engineers; improve data quality and latency from ingestion to insight.
Prototype quick models (from baselines to gradient boosting) to improve matching and scoring.
Help evaluate LLM-powered agents: design rubrics, human-in-the-loop studies, and guardrail canaries.
Skills: data science,analytics,design,python,sql,dbt,models,llm
#J-18808-Ljbffr
0–2 years in data science/analytics or similar; BS/BA in a quantitative field (or equivalent work).
Strong SQL; Python for analysis; comfort with experiment design and causal thinking.
Communicates crisply with engineers, PMs, and leadership; turns analysis into action.
Nice-to-haves: dbt, dashboarding (Hex/Mode/Looker), marketplace or search/recommendation metrics, LLM/agent evaluation.
What you’ll do In Your First Year You’ll Ship Analyses And Experiments That Move Core Product Metrics—match Quality, Time-to-hire, Candidate Experience, And Revenue. You’ll
Define north-star and feature-level metrics for our ranking, interview analytics, and payouts systems.
Design/run A/B tests and quasi-experiments; turn results into product decisions the same week.
Build source-of-truth dashboards and lightweight data models so teams can self-serve answers.
Instrument events with engineers; improve data quality and latency from ingestion to insight.
Prototype quick models (from baselines to gradient boosting) to improve matching and scoring.
Help evaluate LLM-powered agents: design rubrics, human-in-the-loop studies, and guardrail canaries.
Skills: data science,analytics,design,python,sql,dbt,models,llm
#J-18808-Ljbffr