CashNews.co
Data analytics company EXL has introduced the EXL Insurance LLM, an industry-specific language model designed to enhance the efficiency of critical insurance tasks such as claims reconciliation and data interpretation in the insurance sector.
Developed in collaboration with NVIDIA AI Enterprise, the EXL Insurance LLM can support critical claims and underwriting-related tasks including data extraction and interpretation, claims reconciliation, question-answering, anomaly detection and chronology summarisation.
Unlike generic language models, this specialised solution is fine-tuned with proprietary insurance data and an understanding of business process operations, and aims to mitigate common issues such as high indemnity costs and increased compliance risks.
In comparative studies, the EXL Insurance LLM outperformed leading pre-trained models, achieving a 30% increase in task accuracy.
It utilises NVIDIA’s full-stack AI platform, with training and deployment facilitated by the NVIDIA NeMo end-to-end platform.
The advanced training process includes low-rank adaptation and supervised fine-tuning, ensuring the model’s precision in handling insurance-specific tasks.
The system’s capabilities are bolstered by NVIDIA Triton Inference Server, which maximises graphics processing unit efficiency for strong performance.
The EXL Insurance LLM also features NVIDIA NeMo Guardrails, enhancing the user experience with improved input and output management.
Furthermore, the EXL Insurance LLM supports a variety of tasks crucial to the insurance industry, such as processing structured and unstructured data, contextual classification and triaging, and generating insights for customer engagement.
These functions are essential for streamlining operations, from claims adjudication to customer service.
“EXL launches industry-specific insurance Large Language Model” was originally created and published by Life Insurance International, a GlobalData owned brand.
The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site.