Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Discover how Epsilon Data Management successfully migrated their AI-powered marketing identity solution from multi-petabyte on-premises Hadoop and data warehouse systems to a unified Databricks Lakehouse platform in this 39-minute conference talk. Learn about the technical challenges and solutions involved in transitioning massive data volumes while maintaining performance and enabling new cloud-based AI research capabilities within budget and timeline constraints. Explore how Delta Lake, Unity Catalog, MLflow, and LLM endpoints were leveraged to reduce data duplication, improve lineage visibility, accelerate data science workflows, and enable immediate data availability across the entire Epsilon platform while maintaining privacy-safe operations. Understand the architectural decisions and implementation strategies that allowed Epsilon to scale their Decision Sciences solution and deploy marketing solutions across multiple cloud providers and regions for global customers. Gain insights into building AI and data science capabilities at internet scale, with practical examples of how the Databricks platform serves as the foundation for enterprise-level marketing technology solutions. The presentation is delivered by Boaz Super, Vice President of Decision Sciences, and Gairik Chakraborty, Vice President of Database at Epsilon Data Management, providing real-world expertise on large-scale data platform migrations and AI implementation strategies.
