超越 LLM 包装器：面向可靠金融 AI 的上下文工程与三引擎架构
Beyond LLM Wrappers: Context Engineering and the Three-Engine Architecture for Reliable Financial AI
AI 导读
一篇 arXiv 论文提出将金融知识图谱、确定性优化引擎与多智能体 LLM 层整合的三引擎架构，并通过“上下文工程”协调 ERP 系统与多租户的查询上下文。基于约 11 个月、15 次部署的观察，单智能体流水线幻觉率约 7-9%，引入图谱支持的多智能体校验后降至 1% 以下，部署时间从约六周缩短到两周。作者指出 ERP 数据质量与租户模式差异仍是主要挑战，并提出评测、可复现性与隐私等研究方向。
正文 · 原文
Abstract:Financial AI systems must support decision-bearing questions involving payments, collections, cash flow, and compliance while ensuring numerical accuracy, traceability, and auditability. Large language models (LLMs) enable natural language interaction but cannot independently guarantee reproducible calculations or evidence-grounded recommendations. We present a production-grounded architecture integrating three components: a financial knowledge graph for enterprise relationships, temporal context, and provenance; a deterministic optimization engine for financial calculations and decisions; and a multi-agent LLM layer for intent interpretation, task orchestration, explanation, and claim validation. A coordination layer, termed context engineering, selects, ranks, composes, and validates query-specific context across ERP systems, tools, and tenants.
Based on observations from 15 deployments over approximately 11 months, single-agent pipelines exhibited hallucination rates of about 7-9% (5-10% across query types), including invented invoices and unsupported financial claims. Graph-backed multi-agent validation reduced observed rates to below 1%, although these findings are observational rather than causal. ERP data quality and tenant-specific schemas remained major implementation challenges. Reusable mappings, graph templates, and validation policies coincided with a reduction in deployment time from roughly six to two weeks.
We argue that financial context engineering extends beyond text-centric retrieval-augmented generation by requiring structured relationships, temporal validity, exact numerical evidence, access controls, and validation against source records and solver outputs. We outline research challenges in evaluation, reproducibility, privacy, context quality, and cross-tenant generalization.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.39230 [cs.AI] |
| (or arXiv:2610.39230v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.39230 arXiv-issued DOI via DataCite |
Submission history
From: Pradeep Kumar Gs Mr [view email]
[v1]
Thu, 29 Oct 2026 02:50:00 UTC (824 KB)
来源：arXiv cs.AI · arxiv.org