Agentic AI: Insurance Claim Processing System
Article Sidebar
Main Article Content
Traditional insurance claim processing is a manual, labor-intensive function prone to human error and significant delays of 24–48 hours per claim. This paper presents an Agentic AI Insurance Claim Processing System that utilizes a collaborative multi-agent architecture to automate the end-to-end verification pipeline. Powered by the LLaMA 3.1 (8B) large language model via the Groq API, the system orchestrates four specialized agents—Policy, Fraud, Eligibility, and Decision Agents—to validate claims against structured Excel datasets. Experimental results on a curated test dataset of 30 insurance claims demonstrate a 93.33% accuracy rate with a mean processing time of 78 seconds, well within the 2-minute operational threshold, demonstrating the viability of Agentic AI in real-world financial services automation. The per-agent reasoning logs were independently assessed by insurance domain experts and confirmed to meet regulatory audit sufficiency standards, directly addressing the critical explainability gap identified in prior literature.
Downloads
References
S. Dash and A. Sharma, "Artificial Intelligence in the Insurance Industry: A Review of Claim Automation and Fraud Detection," Journal of FinTech Analysis, vol. 4, no. 2, pp. 112–128, 2022.
J. Wu, L. Chen, and M. Patel, "AutoGPT: An Empirical Study of Agentic AI Workflows for Financial Decision Making," ACM Conference on Economics and Computation, pp. 245–261, 2023.
S. Rawat, "Automating Insurance Claim Validation using Large Language Models and Structured JSON Datasets," International Journal of Computer Applications, vol. 186, no. 31, pp. 1–8, 2024.
T. Brown et al., "Language Models are Few-Shot Learners," Advances in Neural Information Processing Systems (NeurIPS), vol. 33, pp. 1877–1901, 2020.
A. Vaswani et al., "Attention Is All You Need," Advances in Neural Information Processing Systems (NeurIPS), vol. 30, pp. 5998–6008, 2017.
R. Zhao and K. Liu, "Fraud Detection in Insurance Claims Using Ensemble Machine Learning Models," Expert Systems with Applications, vol. 195, pp. 116–128, 2022.
P. Kasneci et al., "ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education," Learning and Individual Differences, vol. 103, 2023.
This work is licensed under a Creative Commons Attribution 4.0 International License.
All articles published in our journal are licensed under CC-BY 4.0, which permits authors to retain copyright of their work. This license allows for unrestricted use, sharing, and reproduction of the articles, provided that proper credit is given to the original authors and the source.