About
Insurance COOs and CUOs are proving the impact of AI workflows in weeks, not years. Firemind’s governance-first, AWS-native solutions are already delivering measurable results in underwriting and operations.
| Case / Client | AI workflow & use case | Key results | Time to value & benefits |
|---|---|---|---|
| Industry leader – unstructured data | RAG with Amazon Kendra & Bedrock | 9/10 claim response accuracy; superior summarisation | Faster decision-making |
| MRH Trowe – property report review | GenAI workflow with Textract, SageMaker & Step Functions | 2.5 h → 7 min review time; higher accuracy & repeatability | Delivered in ~4 months; freed account managers |
Industry leader – a real way to sort through unstructured data at scale
Challenge
A leading specialty insurance broker held more than 16 terabytes of unstructured data in its document management system - emails, PDFs, spreadsheets - with limited metadata. Extracting insights was slow and often impossible, leaving valuable knowledge underutilised.
Account managers manually reviewed long property reports, extracting “green” and “yellow” field data into spreadsheets - a 2.5-hour process per document that limited client-facing time.
Firemind implemented a Retrieval Augmented Generation (RAG) workflow using Amazon Kendra and Bedrock, enabling employees to query documents in plain language and get accurate answers.
Firemind designed an automated GenAI workflow orchestrated with AWS Step Functions, Textract, SageMaker, and Pulse. Property data was extracted, structured, and reviewed in minutes, with human oversight before submission.
Results
- 9/10 accuracy on claim file responses
- Claude 2.1 delivered superior summarisation accuracy
- Faster, more informed decision-making across teams
MRH Trowe – cutting property report review from 2.5 hours to 7 minutes
- Reduced review time from 2.5 hours → 7 minutes
- Delivered in just 4 months from first meeting to production
- Higher accuracy and repeatability in data extraction
- Freed account managers for client engagement
Both cases show how AI, applied to specific operational bottlenecks, delivers immediate and measurable gains:
- 75%+ faster processing
- 3× more underwriting/ops capacity
- Proven in production, not just pilots
Firemind combines deep insurance expertise with AWS-native AI and NVIDIA acceleration to deliver operational excellence through proven AI.
