Data Scientist
Zoom · India · 6+ yrs experience · Posted 2026-07-23
Tech stack: Python, SQL
Apply on the company site · Get a referral for this role
Zoom salary & ratings · Zoom interview process · More live openings
About the role
Lead the development of a system converting raw telemetry into actionable insights using automation and scalable MLOps principles within a multi-product SaaS environment. Responsibilities include data exploration, model development, deployment, production monitoring, and incident response. The role focuses on scaling predictive analytics for PQL scoring, expansion modeling, and churn prediction. Collaborate with teams to translate insights into impactful business actions for Sales, Product, and Leadership stakeholders. Deliver a full-lifecycle solution that moves beyond ad hoc analyses to create a robust, production-grade intelligence engine supporting strategic decision-making across the organization. We build data products that power revenue decisions across a multi-product SaaS platform. Our team partners closely with Sales, Product, and Engineering to turn telemetry into action.
Responsibilities:
- Designing and deploying end-to-end ML models — including PQL scoring, churn prediction, and expansion modeling — into production environments with defined latency and accuracy SLAs.
- Building and maintaining MLOps pipelines covering dataset versioning, feature drift monitoring, automated retraining, and model registry management.
- Collaborating with Product and Data Engineering teams to establish telemetry schemas, data contracts, and quality validation frameworks, ensuring models train and score using dependable data.
- Owning model observability by creating dashboards and alerts for performance degradation, prediction drift, and data anomalies — and leading incident response when issues arise.
- Standardising analytics frameworks by developing reusable dbt data models and owning the full experimentation lifecycle, from A/B test design to ship/no-ship recommendations.
Qualifications:
- Demonstrate 6+ years of experience in applied data science or product analytics, or equivalent practical experience.
- Apply advanced SQL and Python (Pandas, Scikit-learn, Statsmodels, PyTorch or TensorFlow) to solve complex analytical problems.
- Build, deploy, and monitor ML models in production environments with a focus on reliability and performance.
- Apply strong foundations in statistics, causal inference, and experimentation design to business problems.
- Work with telemetry and event-level data to model user behaviour and product engagement.
- Use MLOps platforms such as MLflow, SageMaker, Vertex AI, or Kubeflow to manage model lifecycles.
- Apply data quality and observability tools (e.g. Great Expectations, Soda, Monte Carlo) to upstream data pipelines.
- Leverage LLMs or generative AI techniques to automate insight generation from telemetry data.
Qualifications
- Demonstrate 6+ years of experience in applied data science or product analytics, or equivalent practical experience.
- Apply advanced SQL and Python (Pandas, Scikit-learn, Statsmodels, PyTorch or TensorFlow) to solve complex analytical problems.
- Build, deploy, and monitor ML models in production environments with a focus on reliability and performance.
- Apply strong foundations in statistics, causal inference, and experimentation design to business problems.
- Work with telemetry and event-level data to model user behaviour and product engagement.
- Use MLOps platforms such as MLflow, SageMaker, Vertex AI, or Kubeflow to manage model lifecycles.
- Apply data quality and observability tools (e.g. Great Expectations, Soda, Monte Carlo) to upstream data pipelines.
- Leverage LLMs or generative AI techniques to automate insight generation from telemetry data.
Responsibilities
- Designing and deploying end-to-end ML models — including PQL scoring, churn prediction, and expansion modeling — into production environments with defined latency and accuracy SLAs.
- Building and maintaining MLOps pipelines covering dataset versioning, feature drift monitoring, automated retraining, and model registry management.
- Collaborating with Product and Data Engineering teams to establish telemetry schemas, data contracts, and quality validation frameworks, ensuring models train and score using dependable data.
- Owning model observability by creating dashboards and alerts for performance degradation, prediction drift, and data anomalies — and leading incident response when issues arise.
- Standardising analytics frameworks by developing reusable dbt data models and owning the full experimentation lifecycle, from A/B test design to ship/no-ship recommendations.