Contractor - Marketing & Communications

6-9 month contract, open to hybrid and possibly remote arrangements. Candidates should have 5-10 years of experience and a Master's degree, alongside 1-2 years of strong data science and coding experience.

The role focuses on enabling and simplifying analysis required for faster implementation of new CDH modeling features. Key deliverables include:

• Library of required queries/scripts to replicate the CDH customer contextual object in external systems (Databricks/ASL) for deeper analysis.
• Standardizing format for executing key data retrieval steps for use by the broader team, including:
  • Interaction to outcome attribution (account opens).
  • Model data to interaction mapping (model performance, predictor performance).
  • Member Profile to interaction mapping.
• Create notebooks for the broader team to use to answer specific questions, such as:
  • Distribution Analysis.
  • Arbitration Analysis.
  • Channel Engagement Analysis.

Candidates should have familiarity with the Databricks environment and proficiency with Python/PySpark and SQL. Pega CDH experience is preferred.

• Initial Analysis to Support New Model Related Features:
  • Propensity Thresholds:
    • Creating the back-testing approach.
    • Establishing baseline KPIs.
    • Creating the monitoring approach.
  • Initial Model Maturity Analysis:
    • Establishing baseline KPIs.
    • Gauging the impact of enabling the feature.
    • Creating the ongoing monitoring approach.
• On-going Analysis:
  • Model Performance Monitoring: Involvement in creating new logic when changes are needed.
  • NBI Program Model Health: Improving the state of current models for broader sharing.

Addressing analytical gaps that should be readily available:

• "Actionable Monitoring Data": Standardizing analysis for consistency, including:
  • Capturing when propensity scores are exceptionally low closer to real-time (1 day).
  • Capturing when actions are not providing value to their intended objective (acquisition, engagement).
• Eligible Audience Monitoring:
  • Identifying Members eligible for different actions/treatments.
  • Tying interactions back to key Member demographic data for more granular analysis.