Mid Level Data Scientist

Data Scientist

We are seeking a Mid Level Data Scientist to support the automation of media operations workflows. This role involves developing an AI-automated solution to process unstructured audience quality data and building data science models to measure campaign effectiveness. The objective is to create a deployable model with production-grade code that compares on-site user behavior to audience quality metrics, ultimately enhancing how we measure the relevance of advertising campaigns.

Key Responsibilities

• Develop an automated process using AI-native tools to process and load unstructured data from sources like PDFs and emails into Databricks tables.
• Build data science models to analyze user behavior at an aggregate level.
• Create models that compare on-site behavior to audience quality data to determine campaign effectiveness.
• Develop AI agents to extract key fields from unstructured sources, such as audience quality metrics and campaign performance data.
• Standardize, clean, and enrich data by normalizing formats, resolving inconsistencies, and joining datasets.
• Write production-grade code and work towards a deployable model.
• Collaborate with engineering teams on integration with the in-house technology stack.

Required Qualifications

• Experience with Python.
• Demonstrated experience with Agentic AI.
• Familiarity with Large Language Models (LLMs), specifically Bedrock foundation models.
• Experience in data science modeling.
• Proficiency with Databricks or Snowflake.
• Experience working with large datasets (hundreds of gigabytes to terabytes, billions of rows).
• Knowledge of AWS, including Lambdas for orchestration.

Preferred Qualifications

• Familiarity with Document AI or Optical Character Recognition (OCR).
• Experience in data engineering tasks such as unstructured data processing, data cleaning, standardization, and joining/enriching datasets.