SBIR Phase I: Intelligent Financial Coaching Platform Powered by AI and Behavioral Science

About This Grant

The broader/commercial impact of this Small Business Innovation Research (SBIR) Phase I project is to reduce financial literacy gaps among young Americans. Less than 40 percent of this group demonstrate basic financial literacy; most live paycheck to paycheck. This challenge is not only personal but national, contributing to tens of billions in economic loss annually. As this group prepares to inherit trillions in wealth, the stakes grow higher. Traditional financial tools fail to engage young people effectively, pushing them to seek guidance from unreliable sources like social media and unregulated online content. This project addresses that gap through a technologically advanced, personalized interactive financial coaching experience designed for digital-native users. Using applied artificial intelligence, behavioral science, real-time financial data, and natural dialogue, the proposed innovation aims to build better financial habits in a format that aligns with preferred modes of communication. The initial commercialization effort will focus on community banks and credit unions—a $3 billion market. Within three years, the technology aims to improve the financial health of over 250,000 individuals, measured through goal achievement, engagement, and standardized well-being metrics. Positioned within financial technology, the innovation offers a durable competitive advantage through its trust-based, user-driven model. This Small Business Innovation Research (SBIR) Phase I project addresses the technical challenge of replicating human financial coaching through a multi-agent artificial intelligence system that integrates behavioral modeling, natural language processing, and real-time financial data ingestion. The project aims to develop an architecture in which multiple transformer-based agents, specialized for retrieval, personalization, compliance, and planning, work collaboratively within a modular orchestration framework. Using tools such as LangChain for workflow coordination and vector databases for contextual memory retention, the system will deliver adaptive financial guidance tailored to individual user behavior over time. The research will evaluate the extent to which retrieval-augmented generation can support accurate, context-sensitive responses that align with certified coaching practices and behavioral economics principles. Experimental validation will be conducted through structured financial scenarios and iterative testing in focus groups. Technical performance will be measured by agent coordination efficiency, semantic alignment using BLEU and relevance scores, and behavioral impact metrics such as those derived from the Financial Well-Being Scale. Expected outcomes include a functional prototype and empirical evidence supporting the viability of agent collaboration in high-compliance environments. This project contributes to the advancement of applied artificial intelligence in real-time decision support systems within the financial technology domain. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.