Hutch Underwriting is an independent Australian digital underwriting agency specialising in SME Construction, Strata, and Professional Indemnity insurance. Positioning themselves as a “broker-first” business, they leverage a digital platform integrated with major exchanges like Ebix Sunrise and Steadfast’s SCTP to reach 95% of the Australian broker market.
Hutch processes a high volume of inbound insurance quote submissions received via unstructured emails and attachments. This workflow relied on manual review to extract data and re-enter it into their core underwriting system, Entsia. This manual process created significant bottlenecks, resulting in slower response times for brokers and limiting the company’s ability to scale operations efficiently.
While a previous Proof of Concept (PoC) had validated the feasibility of Generative AI, Hutch needed to transition to a production-grade pipeline. The primary technical challenge was the high variability in broker-submitted documents —including nested emails, PDFs, images, and spreadsheets— which required intelligent extraction and strict, deterministic validation against Hutch’s “Question Set” schema to ensure data quality. A failure to automate this reliably would perpetuate operational overhead and prevent Hutch from meeting their goal of “Better Cover” through rapid digital delivery.
DNX, an AWS Premier Consulting Partner, delivered a production-ready, event-driven Generative AI pipeline built on Databricks and AWS. The solution processes inbound emails received via Amazon SES and stored in Amazon S3, utilising Amazon Textract for Optical Character Recognition (OCR) to extract text from PDFs, images and scanned documents.
The core intelligence is powered by Amazon Bedrock using Anthropic’s Claude Sonnet 4 foundation model. DNX implemented a sophisticated prompt engineering approach that transforms unstructured email content into schema-compliant JSON payloads. The system includes an AI-generated confidence scoring mechanism that flags low-confidence extractions for human review, ensuring data quality while maximising automation.
Databricks serves as the orchestration layer, with notebooks managing the end-to-end workflow: email parsing, document extraction, AI processing, schema validation, and API integration. Processed quotes are stored in Delta Lake tables and programmatically submitted to the Entsia Production API and Sunrise for underwriting.
The implementation successfully achieved automated end-to-end submission of quotes into Hutch’s underwriting system. Unstructured emails are now automatically transformed into schema-compliant JSON payloads in near real-time, with AI-generated confidence scores enabling targeted human review of complex submissions. This intelligent quality control ensures data accuracy while maximising automation throughput.
By eliminating manual data entry, underwriters can now focus on complex risk assessment rather than administrative tasks. The Databricks-based architecture provides elastic compute capacity to handle fluctuating quote volumes, while comprehensive logging across all processing stages delivers a complete audit trail for compliance and debugging purposes.
Key outcomes include:
DNX is a systems integrator and trusted advisor for business transformation, focused on democratising modernisation. For us, modernisation is not just about upgrading technology. It is about enabling organisations to adopt modern platforms, architectures and ways of working that drive measurable business outcomes. This includes cloud-native technologies, security and compliance by design, data-driven decision-making, and governance models that support scale, agility and innovation. By democratising modernisation, DNX ensures these capabilities are accessible, practical and aligned to each organisation’s maturity—empowering teams across the business, not just IT, to modernise with confidence and purpose.
The solution leverages the AWS and Databricks services.
