Data Science & ML for Marketing

Data Scientist · OMD EMEA

July 2021 – February 2023

Overview

Led data science initiatives at OMD EMEA, focusing on developing machine learning solutions for marketing and advertising campaigns. This role involved building predictive models, optimizing campaign performance, and modernizing legacy systems to leverage cloud-based ML platforms.

The work combined advanced machine learning techniques with practical business applications, delivering measurable improvements in campaign ROI and operational efficiency for advertising clients.

Key Achievements

Campaign Planning Web Application

Led development of a web application leveraging survey data and XGBoost Shapley values for automated campaign planning. The system optimized ML logic and BigQuery architecture, resulting in dramatic performance improvements.

  • 90% reduction in query times through optimized BigQuery architecture
  • 50% reduction in cloud costs through efficient data processing
  • XGBoost model with Shapley value explanations for transparency
  • Interactive web interface for campaign planning and analysis

200% ROI Increase Through Audience Segmentation

Drove ML roadmap definition with external stakeholders, delivering clustering-based audience segmentation that increased advertising ROI by 200%. The solution identified high-value customer segments and optimized targeting strategies.

  • Advanced clustering algorithms for customer segmentation
  • Real-time targeting optimization based on segment characteristics
  • Integration with advertising platforms for automated campaign execution
  • Continuous monitoring and refinement of segmentation models

Legacy Model Modernization

Modernized legacy models via Vertex AI, improving average accuracy by 5% while reducing maintenance overhead. The migration to cloud-based ML platforms enabled scalable, production-ready solutions.

  • Migration from on-premises to Vertex AI cloud platform
  • Model retraining with improved feature engineering
  • Automated model deployment and monitoring pipelines
  • Reduced operational costs through cloud infrastructure

ML Roadmap & Strategy

Collaborated with external stakeholders to define ML roadmap, identifying opportunities for AI-driven improvements in marketing effectiveness. The strategy balanced technical innovation with business objectives, ensuring alignment with client goals.

  • Strategic planning sessions with client stakeholders
  • Use case identification and prioritization
  • Technical feasibility assessments and proof of concepts
  • Resource planning and timeline estimation

Technologies & Tools

XGBoost BigQuery Vertex AI Clustering Python GCP Shapley Values Web Development

Impact & Results

90%

Query time reduction

200%

ROI increase

5%

Accuracy improvement

Project Highlights

Campaign Planning Platform

Web-based application for automated campaign planning using ML predictions and Shapley value explanations for transparency and trust.

XGBoost BigQuery Web App

Audience Segmentation System

Clustering-based system for identifying high-value customer segments and optimizing advertising targeting strategies.

Clustering ML GCP
Back to Experience