Natural Language-Driven Data Visualization Using Kestrel AI: A Novel Algorithm for Intelligent Analytics

Generative AI in financial analytics is increasingly vital for interpreting vast, complex datasets to guide strategic decisions in dynamic, real-time markets, yet its technical complexity often limits non-expert engagement. Traditional methods, rooted in structured query language (SQL), depend heavily on specialized expertise, restricting access for business analysts and decision-makers lacking programming skills. These conventional systems achieve 75 percent to 80 percent accuracy for simple queries but falter significantly with complex, nested conditions or multi-operator logic, diminishing their utility in today’s fast-paced financial landscape. We present Kestrel AI, a pioneering algorithm harnessing advanced Natural Language Processing (NLP) and Retrieval-Augmented Generation (RAG) to convert natural language queries into precise SQL commands and insightful visualizations. With a modular, scalable architecture and HighChart integration, Kestrel AI delivers an outstanding 92 percent average accuracy, substantially outperforming traditional approaches across diverse query types, from basic to intricate. Experimental validation shows query execution times averaging 0.3 seconds, highlighting its speed and efficiency, while user studies reveal 85 percent of participants, spanning seasoned financial analysts to complete novices, commend its intuitive interface and rapid insight delivery. The system adeptly processes structured and unstructured data, allowing users to blend diverse datasets for thorough analysis, and adapts seamlessly to real-time market shifts. By simplifying complex data interactions and enhancing access, Kestrel AI fosters inclusive, data-driven decision-making in financial contexts. This work sets a new standard for generative AI-driven analytics, offering a transformative tool for investment research, market analysis, and organizational collaboration that bridges technical barriers and user-friendly design.