Lead Data Scientist

AT&T AT&T · Telecom · Atlanta, GA +1

Lead Data Scientist role focused on developing, deploying, and optimizing machine learning survival models to predict customer tenure for AT&T Fiber and AT&T Internet Air customers. This involves end-to-end data science workflows, from data extraction and feature engineering to model tuning, interpretation, and deployment, with a focus on informing strategic business decisions and estimating customer lifetime value.

What you'd actually do

  1. Lead the development, deployment, and optimization of survival models for broadband, AT&T Fiber, and AT&T Internet Air customers to support customer lifetime value estimation and strategic business decisions.
  2. Translate complex business problems into actionable insights through end-to-end data science workflows, including data extraction, cleansing, feature engineering, exploratory data analysis, model development, tuning, interpretation, deployment, and ongoing model monitoring.
  3. Design, build, and analyze large, complex data sets from structured and unstructured sources, including data lakes, databases, cloud platforms, and enterprise data environments, while ensuring data quality, integrity, and usability.
  4. Use simulation techniques and statistical analysis to identify key drivers of customer survival, interpret model outputs, and deliver clear, actionable recommendations to business stakeholders.
  5. Develop scalable coding solutions using Python and PySpark/Spark, with proficiency in modern machine learning libraries and frameworks to support robust model development and production-ready implementation.

Skills

Required

  • Python
  • scikit-learn
  • pandas
  • NumPy
  • PySpark/Spark
  • Databricks ML
  • SQL
  • Databricks
  • Snowflake
  • survival modeling techniques
  • machine learning methods
  • applied statistical analysis
  • data visualization
  • storytelling
  • communication skills
  • MLflow

Nice to have

  • data lakes
  • cloud platforms
  • enterprise data environments

What the JD emphasized

  • survival modeling techniques
  • machine learning methods
  • applied statistical analysis
  • Python
  • PySpark/Spark
  • Databricks ML
  • SQL
  • Databricks
  • Snowflake
  • MLflow

Other signals

  • develop, deploy, and optimize machine learning models
  • survival models
  • customer lifetime value
  • translate business problems into actionable insights
  • end-to-end data science workflows