Sr. Data Scientist with Statistics Background - Twitch

Required Skills

• Statistics Time-Series Analysis: EWMA, seasonal decomposition, statistical process control, change-point detection, hypothesis testing, confidence intervals. Ability to design methods that generalize across new metric categories without per-category tuning.
• Python: Strong proficiency in Python with experience building production-quality data processing and automation solutions
• SQL: Advanced querying on large-scale warehouses (Redshift or equivalent) - window functions, CTEs, complex joins, aggregations across multi-table schemas with millions of rows.
• NLP / Text Analytics: Clustering, topic modeling, or embedding-based approaches for identifying emerging categories from unstructured text (chat transcripts, support cases).
• Data Integration: Comfortable pulling and correlating data across heterogeneous sources (data warehouses, Jira, Salesforce, Google Calendar, Slack).

Nice to Have

• Experience building alerting/monitoring systems or operational dashboards
• Familiarity with CX/support metrics (CSAT, resolution rates, escalation rates, case volumes)
• Root cause analysis frameworks - correlating metric anomalies to upstream events (deployments, content changes, feature rollouts)
• Experience with configurable/extensible system design (pluggable analyzers, threshold tuning, multi-source ingestion)

Required Skills

• Statistics Time-Series Analysis: EWMA, seasonal decomposition, statistical process control, change-point detection, hypothesis testing, confidence intervals. Ability to design methods that generalize across new metric categories without per-category tuning.
• Python: Strong proficiency in Python with experience building production-quality data processing and automation solutions
• SQL: Advanced querying on large-scale warehouses (Redshift or equivalent) - window functions, CTEs, complex joins, aggregations across multi-table schemas with millions of rows.
• NLP / Text Analytics: Clustering, topic modeling, or embedding-based approaches for identifying emerging categories from unstructured text (chat transcripts, support cases).
• Data Integration: Comfortable pulling and correlating data across heterogeneous sources (data warehouses, Jira, Salesforce, Google Calendar, Slack).

Nice to Have

• Experience building alerting/monitoring systems or operational dashboards
• Familiarity with CX/support metrics (CSAT, resolution rates, escalation rates, case volumes)
• Root cause analysis frameworks - correlating metric anomalies to upstream events (deployments, content changes, feature rollouts)
• Experience with configurable/extensible system design (pluggable analyzers, threshold tuning, multi-source ingestion)