NeuralMetrics Launches Platform Using AI Agents to Enhance Insurance Underwriting

The A-Star Platform’s ‘SAMA’ autonomous AI agents can learn, adapt, and self-correct instantaneously to set new standards in data-driven risk analysis for commercial underwriting.

(Image source: NeuralMetrics website.)

NeuralMetrics (Denver), a provider of AI-generated commercial underwriting data, has announced the launch of its A-Star (A* informed search/pathfinding) AI platform, consisting of intelligent AI persona agents that serve to expand the automation of commercial underwriting workflows. The persona-based AI platform delivers accurate responses to risk questions and automates complex risk-assessment tasks without compromising data source transparency and regulatory compliance, according to a statement from the vendor. Versatile AI agents or knowledge assistants can assume distinct personas, such as risk-assessment agents or document/content classification agents. Referred to as Smart Adaptive Multifunctional Systems Agents (SAMAs), the AI assistants excel at real-time learning, adapt to their tasks, and can be recalibrated or self-calibrate for continuous optimal performance in support of data-driven underwriting, the statement adds.

Marcus Daley, Co-Founder, NeuralMetrics.

The NeuralMetrics A-Star seeks to realize the potential of Massive Multitask Language Understanding (MMLUs), derived from advances in Large Language Models (LLMs), and then applied to the challenges associated with timely data acquisition for risk assessment. The platform generates synthetic data and triangulates it with a wide range of real-world data, allowing the AI agents to learn and choose the best answers to risk questions, according to the vendor. Additionally, the platform’s GenAI capabilities facilitate AI agents to trace back and auto-correct output when initial responses to risk questions require modification.

The A-Star platform is the first of several initiatives resulting from joint research between NeuralMetrics and Binghamton University. Focusing on the use of quantum computing and AI, the collaboration aims to discover other innovative methods to generate synthetic data and advance the proficiency of autonomous AI agents. The research and outcomes are significant leaps forward for the underwriting data products and capabilities of NeuralMetrics, with additional advances underway, the vendor asserts.

“Pushing the boundaries of artificial intelligence, the NeuralMetrics A-Star AI platform enables real-time insights about risk and exposures, while ensuring strong compliance and transparency in the use of diverse, up-to-the-minute risk-quality data elements,” comments Marcus Daley, co-founder, NeuralMetrics. “Initial implementation in commercial underwriting demonstrates a notable increase in predictive accuracy, helping our innovation journey to approach a 100 percent risk data recall rate. Additionally, advancements in synthetic data generation and AI agent reasoning open new opportunities in vertical industries for NeuralMetrics, enabling the company to expand into new market segments and business domains.”

SAMAs Underpin Existing Underwriting Data Products

NeuralMetrics says that the first set of Smart Adaptive Multifunctional Systems Agents (SAMAs) underpins the vendor’s existing underwriting data products, improving the its abilities to provide immediate risk insights and enhance exposure management in commercial underwriting. NeuralMetrics says it is continuing AI agent development to increase risk-evaluation productivity, as autonomous assistant personas are perpetually augmented with numerous combinations of real-world and synthetic data.

NeuralMetrics Launches Generative AI Underwriting Data Workbench

Anthony R. O’Donnell // Anthony O'Donnell is Executive Editor of Insurance Innovation Reporter. For nearly two decades, he has been an observer and commentator on the use of information technology in the insurance industry, following industry trends and writing about the use of IT across all sectors of the insurance industry. He can be reached at or (503) 936-2803.

Leave a Comment