Case Study: Convex speeds specialty insurance underwriting with Provectus GenAI document intelligence

Convex cuts 100-page risk reports to 10 minutes with Provectus

Convex, a specialty insurance company, faced a bottleneck in its underwriting process where subject-matter experts spent hours manually reading and summarizing lengthy, complex risk reports. This manual work slowed down decision-making and limited the growth of their portfolio. To address this, Convex partnered with Provectus to implement a generative AI document intelligence solution.

Provectus built a production-ready GenAI solution using a tuned large language model on Amazon Bedrock. The system automates the extraction, segmentation, classification, and summarization of 100-page reports into structured data. This reduced the summarization time from hours to just ten minutes, allowing Convex to triage 100 times more risk reports against the same book and significantly grow their portfolio.

Convex