Democratizing Data for AI: Varo Bank's Approach to Scaling AI Innovation

In fintech, data is the cornerstone of innovation and customer-centric AI solutions. In this technical deep dive, Tim Hopper, Staff Software Engineer, Machine Learning on Varo Bank's ML platform team, presents how Varo streamlined their approach to productionizing data, driving AI/ML initiatives across multiple products and consumer experiences.

Tim will delve into Varo's journey of democratizing data transformations, empowering data scientists, engineers, and analysts to collaboratively define and share features. This approach has not only accelerated machine learning model development but also opened up new avenues for data utilization across the organization.

Tim will also showcase how Varo has implemented a unified feature platform that seamlessly integrates batch and streaming data, serving both online and offline use cases. He’ll discuss the unexpected benefits of this approach, including the ability to repurpose data transformations for multiple applications such as rules engines, customer-facing dashboards, and future machine learning models.

• Technical challenges and solutions for unifying batch and streaming data pipelines
• Strategies for extending real-time data serving beyond ML to operational use cases
• Maximizing feature reuse to streamline production and reduce costs
• Architectural insights on leveraging a unified feature platform that serves both online and offline stores
• Best practices for empowering cross-functional teams to contribute to and use a centralized feature platform