Title

AWS re:Invent 2023 - How Rocket Companies run their data science platform on AWS (ANT321)

Summary

• Introduction: Ravindra Gupta, the worldwide go-to market lead for machine learning at AWS, introduces the session on leveraging AWS analytics and machine learning services to harvest data.
• Rocket Companies' Journey: Dian Xu from Rocket Companies shares the company's journey, from its founding principles to its evolution into a provider of a wide range of financial services.
• Data Challenges and Solutions: Rocket Companies faced challenges with their legacy data infrastructure, which was expensive, not scalable, and difficult to support. They modernized their data platform using AWS services, resulting in faster data ingestion, reduced support issues, and cost savings.
• Modernized Data Science Platform: The new platform includes AWS services like S3, Lake Formation, Glue Catalog, Lambda, EMR, Redshift, and SageMaker. This modernization has led to significant improvements in data processing and analytics capabilities.
• Gen AI Readiness: Rocket Companies is ready for the next generation of AI, having joined an early preview for AWS Bedrock and providing access to large language models (LLMs) to their data scientists.
• Future Plans: Rocket Companies aims to continue enhancing data governance and leveraging Gen AI to improve data management and engineering.
• AWS Features and Capabilities: The session concludes with an overview of AWS features that support data science and machine learning, including SageMaker Canvas, Studio, MLOps capabilities, and EMR Serverless.

Insights

• Data as a Critical Asset: The session underscores the importance of data in modern business and the need for effective tools and strategies to harvest and utilize it.
• Cost-Effectiveness and Scalability: The transition from legacy systems to AWS services highlights the benefits of cloud-based solutions in terms of cost savings, scalability, and reliability.
• Cultural Emphasis on Innovation: Rocket Companies' culture of innovation and their "ISMs" principles are central to their success in modernizing their data platform and embracing new technologies.
• Empowering Diverse Roles: AWS services like SageMaker Canvas enable non-technical roles to participate in machine learning, democratizing data science across the organization.
• Preparedness for Gen AI: Rocket Companies' investment in a modernized data platform has positioned them to quickly adopt and benefit from the next wave of AI advancements.
• AWS as an Enabler: The session demonstrates how AWS services can be leveraged to build a robust, modern data science platform that meets the needs of various stakeholders, from data scientists to business analysts.