Case Study: Johnson Lambert LLP speeds up insurance audits with Provectus GenAI on Amazon Bedrock

Johnson Lambert LLP cuts financial risk audit time from 80 hours to 12 with Provectus

Johnson Lambert LLP, a CPA and consulting firm for the insurance industry, faced a significant challenge as its auditors were spending 60 to 80 hours per financial risk audit manually processing PDF reports and tracing data. This intensive manual work diverted resources from higher-value client-facing activities. They partnered with vendor Provectus to implement a generative AI solution to address this bottleneck.

Provectus built a document-to-structure pipeline on Amazon Bedrock that uses large language models to extract, normalize, and validate data from unstructured audit reports. The solution reduced the time required for a single financial risk audit from 60-80 hours to just 12-16 hours, a dramatic 50% reduction in document processing time. This led to a 20% efficiency gain across Johnson Lambert's entire audit book, allowing auditors to focus on review and analysis instead of manual data entry.

Johnson Lambert LLP