Associate Data Scientist

Overview

Associate Data Scientist

• on a builder path.
• This role blends product thinking, applied statistics, and production-minded analytics.
• You will:
• turn ambiguous problems into measurable questions
• build reliable measurement (metrics and experimentation)
• create lightweight models and decision tools that teams can use in real workflows
• partner closely with engineering, product, and risk to improve customer outcomes and business performance
• build AI-assisted decision workflows (for example, LLM-powered analysis copilots backed by a curated metrics layer, with clear guardrails and validation)
• With the advancement of AI, we value people who have strong fundamentals and clear thinking.
• Understanding data generation, measurement, tradeoffs, and how to validate results matters more than memorizing tools.
• You'll learn how to use AI responsibly to move faster, while still owning correctness, robustness, and interpretation.
• Your responsibilities
• Define problems and measurement
• Translate product or business questions into testable hypotheses and clear success metrics.
• Build and maintain metric definitions and analysis templates so decisions are consistent and repeatable.
• Experimentation and causal thinking
• Design, analyze, and interpret A/B tests (and quasi-experiments when randomization is not possible).
• Partner with product and engineering to ensure correct tracking, guardrails, and experiment quality.
• Modeling for decisions (practical ML)
• Build baseline predictive and segmentation models (for example, churn propensity, risk signals, customer clustering) with clear evaluation and limitations.
• Focus on models that are actionable : who will use them, when, and what decision they change.
• Applied AI (LLMs) for analytics and decisioning
• Prototype small, safe LLM use cases that improve how teams explore data (for example, natural-language Q&A over a curated dataset, summarizing experiment results, or generating investigation checklists).
• Help define evaluation and validation approaches for AI-assisted outputs (sanity checks, golden queries, offline test sets, and human review loops).
• Partner with platform and security stakeholders to ensure AI workflows are permissioned correctly and avoid leaking sensitive data.
• Analytics that ships
• Create reusable, well-documented analysis assets: datasets, feature tables, notebooks, dashboards, and simple services or jobs.
• Collaborate with data engineering to productionize reliable pipelines where needed.
• Data quality and reliability
• Validate data inputs, monitor key metrics, and investigate anomalies.
• Document assumptions and build sanity checks so analyses and models are trustworthy.
• Responsible use of AI tools
• Use AI to accelerate coding, documentation, and exploration.
• Validate AI outputs, protect sensitive data, and follow safe data handling practices.