Rocket Mortgage leverages GenAI to enhance efficiency, accuracy, and decision-making in the mortgage industry.

Rocket Mortgage, an American mortgage lender and pioneer in using ML and AI, is adopting a model-agnostic approach to GenAI. This strategy, supported by a versatile data platform, emphasizes speed and flexibility. By integrating multiple GenAI models and maintaining a “human in the loop” approach, Rocket Mortgage ensures precise and efficient operations.

One key application is a GenAI chatbot in development, designed to listen to and comprehend queries effectively. Another assistant verifies employer identities, speeding up the decision-making process by understanding various names for the same company, such as Google and Alphabet. This approach is more efficient than traditional database methods.

Rocket Mortgage has implemented several business processes with GenAI, all while maintaining proper guardrails to assure investors and regulators of the technology’s safety and responsibility. “With genAI-powered copilots or systems, which is a lot of what we’re building, we find that, with the combination of a genAI model that knows everything posted on the internet for years, and human judgment, the accuracy of the decision is going to increase 10% to 15%, which is huge,” says CIO Brian Woodring.

The company uses AWS Bedrock and Anthropic AI for its GenAI models, maintaining a model-agnostic strategy to choose the right tool for each task. Rocket Mortgage’s modern data platform supports structured, semi-structured, and unstructured data, making it highly accessible for AI models. This data-centric approach ensures that AI applications can rapidly access and utilize the necessary information for efficient mortgage processing.