[提交于 2026年7月8日]

Agentic AI与检索增强模型在直通式核保中的应用

查看题为《Agentic AI and Retrieval-Augmented Models in Straight-Through Underwriting》的论文 PDF，作者 Robert Richardson 及三位其他作者

查看 PDF HTML（实验性）

摘要：人工智能（AI）正开始重塑精算实践，特别是在需要对非结构化文档、异构数据源和受监管的决策工作流进行推理的领域。精算师如今面临的设计空间涵盖从传统的基于规则的自动化到大语言模型（LLM）、检索增强生成（RAG）以及能够规划、检索、调用工具并反思的“代理型”多智能体系统。本文探讨了这些新兴架构如何支持精算的优先事项，如透明性、可审计性以及人在环路治理，并以直通式决策流程为重点。为使这些想法具体化，我们开发并分析了一个针对小型商业企业主保单（BOP）直通式核保的代理型AI框架。我们构建了一个合成但贴近实际的实验环境，并比较了三种核保流水线：（i）单一LLM基线，（ii）朴素RAG系统，以及（iii）结合了定向检索、第三方数据校验和明确的多步骤规则评估的多智能体“代理型RAG”流水线。代理型系统整体表现最佳，在多步骤和缺失信息场景中优势最为显著，在这些场景中，结构化检索与反思帮助模型避免了无根据的直通式决策。

提交历史

来自: Robert Richardson 查看邮箱

[v1] 2026年7月8日周三 18:43:34 UTC (3,586 KB)

[Submitted on 8 Jul 2026]

Title:Agentic AI and Retrieval-Augmented Models in Straight-Through Underwriting

View a PDF of the paper titled Agentic AI and Retrieval-Augmented Models in Straight-Through Underwriting, by Robert Richardson and 3 other authors

View PDF HTML (experimental)

Abstract:Artificial intelligence (AI) is beginning to reshape actuarial practice, particularly in domains that require reasoning over unstructured documents, heterogeneous data sources, and regulated decision workflows. Actuaries now face a design space that ranges from traditional rule-based automation to large language models (LLMs), retrieval-augmented generation (RAG), and multi-agent agentic'' systems that plan, retrieve, call tools, and reflect. This paper examines how these emerging architectures can support actuarial priorities such as transparency, auditability, and human-in-the-loop governance, with a focus on straight-through decision processes. To make these ideas concrete, we develop and analyze an agentic AI framework for straight-through underwriting of small commercial Business Owner Policies (BOPs). We construct a synthetic but realistic experimental environment and compare three underwriting pipelines: (i) a single-LLM baseline, (ii) a naive RAG system, and (iii) a multi-agent Agentic RAG'' pipeline that combines targeted retrieval, third-party data checks, and explicit multi-step rule evaluation. The agentic system performs best overall, with the largest gains in multi-step and missing-information scenarios, where structured retrieval and reflection help the model avoid unsupported straight-through decisions.

Submission history

From: Robert Richardson [

view email

]

[v1]

Wed, 8 Jul 2026 18:43:34 UTC (3,586 KB)