Knowledge Management AI is a web application developed using Next.js, React, and TypeScript. This project allows users to interact with a local Language Model (LLM) through chats and conversations based on pre-loaded documents. The application supports a variety of file formats, including PDF, DOCX, and MP3, and at the same time offers a flexible document management.

Use ChromaDB for efficient vector database management and Ollama for local AI model execution. Users can upload and organize documents with a tagging system, with the option for data obscuration in the vector database to ensure privacy. A key feature is the ability to initiate multiple chat sessions on uploaded documents, with the flexibility to change AI models for in the middle of a conversations. The user interface is customizable with various themes, and users have full control over session management, including the option to delete them.

Working on Knowledge Management AI taught me how to integrate complex AI models into a web application and manage large amounts of data with database solutions. It improve my understanding of software architecture patterns . Additionally, I gained hands-on experience in implementing file management systems and building customizable user interfaces with modern web technologies.

Full details on system requirements and build instructions are available on the project’s GitHub page.