Case Study: Mars Petcare achieves predictive, AI-driven pet healthcare with Databricks

Healthier, happier pets with the help of data and AI

Mars Petcare, a division of Mars with more than 50 pet-focused brands, set out to create a holistic view of animal health from diverse data sources—veterinarian notes, dietary records, genomics and more. Progress was hampered by siloed teams and inconsistent data formats: each brand had its own systems, ETL processes and models, which made building pipelines slow and limited collaboration across analytics and data science teams.

To solve this, Mars Petcare built a Petcare Data Platform on Azure Databricks and Delta Lake, unifying data access via JDBC, providing ACID-compliant, versioned datasets, and enabling collaborative notebooks for cross-team work. The platform simplified ETL, sped time-to-insight, lowered cloud costs through autoscaling and easier infrastructure management, and empowered teams to develop predictive models—improving diagnosis and prediction of pet health issues, including terminal diseases, behavioral indicators and genetic conditions.

Mars Petcare