Case Study: Experian boosts customer satisfaction with Databricks

Experian scales customer support with GenAI-powered email automation

Experian, the multinational data broker and consumer credit reporting company, wanted to improve its GenAI initiatives and customer support efficiency as inquiry volumes grew and response times slowed. Working with Databricks, the company set out to build an AI chatbot for contact center emails, aiming to maintain high-quality service while scaling support operations.

Databricks helped Experian implement “Latte” using Mosaic AI, Vector Search, Model Serving, Unity Catalog, and MLflow to fine-tune, deploy, and govern the chatbot end to end. The solution improved retrieval quality, sped up responses, and reduced manual workload, while Databricks delivered up to 6x initial token throughput at the same hourly cost and helped Experian raise customer NPS by 8%.

Experian

James Lin

Head of AI/ML Innovation