Operates as an intelligent CRM backend built on the Model Context Protocol, using vector embeddings and semantic search to enhance customer relationship management. It provides robust user management capabilities, including the creation, retrieval, and semantic searching of user profiles. A core feature is its ability to recommend personalized offers, such as PicPay financial products, based on a user's semantic profile and current financial goals, incorporating risk assessment. Designed with a modular architecture, it supports multiple database types (SQLite and PostgreSQL) and leverages FAISS for efficient vector similarity search, all while ensuring structured logging and comprehensive email validation.
主要功能
010 GitHub stars
02Support for multiple databases (SQLite and PostgreSQL)
03Robust email validation and structured logging
04FAISS-powered vector store for efficient similarity search
05User management with vector embeddings for semantic search
06Intelligent offer recommendations based on semantic profile and goals
