New AI Features to Raise the Bar on Underwriting Efficiency
With our latest product updates, we’ve extended our commercial underwriting product with a suite of AI-powered features to facilitate end-to-end underwriting across all lines of business, scaling from transactional underwriting to complex, three-dimensional risks.
Sixfold’s number one superpower is to easily–and quickly–ingest carriers’ unique underwriting guidelines and automatically surface the submissions that match the carrier’s unique risk appetite. The platform empowers carriers to streamline the underwriting process by:
🔘 Analyzing publicly available information and ingesting data from multiple disparate sources in an instant for a comprehensive risk assessment
🔘 Generating a comprehensive summarization of the business’s operations and providing NAICS/SIC classification
🔘 Surfacing positive and negative risk factors tuned to a carrier’s unique appetite
🔘 Answering complex questions across large sets of documents
🔘 Prioritizing risks with an underwriter-facing dashboard for improved resource allocation
Over the past half, we’ve significantly matured our P&C underwriting platform with deep investments across accuracy, traceability, performance, and extensibility. Some of the key highlights include:
✔️ Improved Accuracy with Advanced Document Extraction
As underwriters ask complex questions across large sets of documents, we’ve invested in new models to improve extraction across the universe of insurance documents. With Sixfold’s latest models, we’ve seen a 40% boost in accuracy in extracting data from the most illegible documents.
With our ongoing investments in extraction models tailored to the documents underwriters see daily, Sixfold ensures precise and reliable insights from large, challenging document sets, transforming how underwriters interact with data.
✔️ Appetite Match Scoring with Weighted Risk Scoring
To replicate underwriters’ cognitive processes, Sixfold is introducing weighted risk signals to reflect the nuance across how underwriting factors impact where a risk sits within a carrier’s appetite. Now, carriers can assign varying importance to different factors to prioritize risks more accurately based on alignment with their risk appetite.
✔️ Enhanced Transparency with Inline Citations
Not only can Sixfold answer complex questionnaires across lines of business, but now all answers are grounded in the original source material with citations to the most relevant sections for confident decision-making.
✔️ Bringing Risk Classification and Summarization Down to Seconds
With our continued commitment to investing in the Sixfold pipeline architecture to improve performance, availability, and resilience, we’ve brought down the median case processing time from 80 seconds to 31 seconds. With these enhancements, in less than a minute, Sixfold can research publicly available information to learn everything we can about a business and analyze the aggregated data for business summarization and NAICS/SIC classification.
✔️ Embed Sixfold across Underwriters’ Existing Workflows with our API
With the Sixfold API, carriers can seamlessly integrate Sixfold into existing workflows for enhanced productivity and unified risk management. From automated data gathering and ingestion to custom-tailored underwriting recommendations that can be embedded across existing workflows and systems, Sixfold cuts out the manual work and document handling overhead for 10x faster risk review processes.
✔️ Mitigating drift and bias with our Responsible AI framework
Designed to navigate the rapidly evolving AI landscape confidently, Sixfold’s Responsible AI framework ensures carriers are well-insulated from risk with its enhanced auditability, data provenance, and traceability. By actively collaborating with regulatory bodies and legal counsel, Sixfold remains at the forefront of responsible AI innovation, safeguarding carriers with unparalleled diligence.
👀 Coming Soon: Research Assistant
Stay tuned for future launch updates to hear about upcoming capabilities like Sixfold’s Research Assistant, designed to find answers to complex research questions, all of which are grounded in the original source material with citations.