Concirrus Launches AI-Native Underwriting Platform
Concirrus has launched Inspire, an AI-native underwriting platform designed for specialty insurers and Managing General Agents (MGAs). The platform aims to accelerate underwriting decisions while maintaining risk controls. It is built to help underwriters handle complex submissions more efficiently.
- The architecture of the Inspire platform combines model orchestration, domain-specific knowledge graphs, and agentic AI to achieve high accuracy and explainability in underwriting processes. This proprietary agentic framework utilizes collaborative agents to process data, challenge findings, and iteratively refine results, which significantly reduces errors in submission ingestion. This approach allows for the automation of complex tasks like submission ingestion, loss run analysis, and portfolio oversight. - A key technical feature is the platform's submission ingestion engine, which uses advanced AI and machine learning to extract, validate, and structure data from complex, often unstructured, submission documents with 95-98% accuracy. This engine is designed with a privacy layer to leverage Large Language Models (LLMs) like Google Bard or ChatGPT securely, preventing data leakage of sensitive information. For example, in a partnership with Applied Aviation, Concirrus's aviation-specific LLM platform reduced risk analysis and quoting time from 36 hours to under five minutes. - The platform is built on a modular, API-first architecture, allowing for incremental adoption and integration with existing core systems, data providers, and rating engines. Modules are interoperable and span the full underwriting lifecycle, including submission intake, triage, quoting, exposure management, data extraction, policy administration, and rating. This design avoids vendor lock-in and the need for repeated re-platforming. - For developers, the platform offers a Python API that wraps the GUI framework, allowing for the creation of custom user interface components and the automation of core functionalities. The API can be used to execute Python scripts in batch mode for backend processes or to extend the platform's user interface. - Concirrus has formed strategic partnerships to extend the platform's capabilities, such as an integration with Diesta for automated premium payment and reconciliation, creating a seamless quote-to-cash workflow. This indicates a strategy of building an ecosystem around the core underwriting platform. - The company has achieved triple certification for ISO/IEC 42001 (AI governance), ISO/IEC 27001 (information security), and SOC 2 compliance, a notable achievement for an insurtech. This focus on governance and security is critical for deploying AI in a regulated industry like insurance. - The emphasis on creating a "golden age for underwriting talent" suggests a human-in-the-loop design philosophy, where the platform automates repetitive tasks to allow underwriters to focus on strategic risk assessment. This approach is demonstrated in a case study with Applied Aviation, where the platform's implementation led to a doubling of gross written premium without a proportional increase in headcount. - The venture capital landscape for AI-native underwriting platforms is robust, with companies like Artificial Labs and Pibit AI raising significant funding rounds. This broader market trend indicates strong investor confidence in the potential for AI to modernize the specialty insurance sector, validating the problem space Concirrus is addressing.
