Authors: Gopichand Talluri
Abstract: The increase in the complexity of financial regulation and the increase in the number of unstructured financial data have presented significant challenges to the conventional compliance systems. The potential solution could be the intelligent automation and processing of financial data via context, since nowadays the development of Large Language Models (LLMs) allows using smart automation and processing of financial information. The paper creates an outline of the operationalization of the LLMs to the financial compliance in managed cloud AI platforms. The proposed system will include pre-processing of the data, inference with the help of the LLM, compliance analysis and continuous monitoring to guarantee the scalability, reliability, and compliance. Experimental analysis of simulated data shows that the proposed model shows better results compared to current methods, including FinBERT, BloombergGPT, and FinGPT in accuracy, efficiency, and compliance score. The findings indicate that the integration of the capabilities of LLM and cloud-based infrastructure is effective in tackling real-world financial compliance issues. This piece of work is a step in the right direction of establishing scalable, reliable, and smart compliance systems in contemporary financial settings.
DOI: https://doi.org/10.5281/zenodo.19732571
