Supporting financial education with a personalized AI chatbot

Our project created an AI-powered financial assistant called Kami, designed specifically for the Ecuadorian context to help with various financial matters, including personal advice, money management, and answering financial questions. Available through a mobile app, Kami makes financial guidance accessible to anyone with a smartphone. The system combines Large Language Models (LLM) with a specialized knowledge base that provides accurate, personalized and relevant information. We host this system on secure cloud infrastructure that protects user privacy while ensuring the service can scale as usage increases.

A central component of the solution is the K-Score, a financial wellbeing index we developed separately from the chatbot technology. The K-Score evaluates three key dimensions: financial reality (income, expenses, and debt), emotional relationship with money (attitudes and feelings toward finances), and financial habits and behaviors (patterns that affect financial health). This comprehensive assessment forms the basis for the guidance provided. The K-Score works together with the AI system, allowing for personalized financial advice based on each user's situation.