Data Scientist

📋 Description

• • Own the full data science lifecycle—from raw clickstream logs in BigQuery to production-grade ML models that predict which video will go viral tomorrow—across a portfolio that reaches 250 million fans every month.
• • Translate petabytes of web, social, and streaming data into clear, monetizable insights for brands like FailArmy, People Are Awesome, and Reader’s Digest, directly influencing editorial calendars, ad-sales packaging, and product roadmaps.
• • Architect and maintain scalable ETL/ELT pipelines on Google Cloud Platform (BigQuery, Dataform, Cloud Functions) that ingest 5B+ events per month, ensuring data quality, lineage, and sub-second query performance for both scheduled and ad-hoc workloads.
• • Design and deploy machine-learning models for traffic forecasting, churn prediction, content-recommendation engines, and dynamic ad-inventory pricing; push models to Vertex AI endpoints with CI/CD guardrails and real-time monitoring for drift and bias.
• • Partner with editors, producers, and growth marketers to run rigorous A/B and multivariate tests—leveraging causal-inference techniques—to prove which thumbnails, headlines, and story angles lift CTR, watch-time, and revenue per session.
• • Build Looker dashboards and self-service data marts that empower non-technical stakeholders to explore KPIs such as CPM uplift, funnel drop-off, and cohort LTV without writing SQL, while you maintain the semantic layer and row-level security.
• • Mine unstructured text and video metadata using NLP and multimodal embeddings to surface trending topics, optimize SEO clusters, and power semantic search experiences that keep users engaged 15 % longer.
• • Champion an experimentation culture by evangelizing best practices in hypothesis framing, sample-size calculation, and post-test analysis; mentor junior analysts and run monthly “data deep-dives” to upskill the broader organization.
• • Continuously scan the horizon for emerging AI/ML tooling—vector databases, LLM agents, reinforcement learning—to automate repetitive workflows (tagging, clipping, thumbnail generation) and unlock new revenue streams.
• • Present concise, visually compelling narratives to the C-suite that translate model outputs into concrete actions, investment asks, and risk assessments, ensuring data science remains a strategic pillar of TMB’s growth story.