MaxKB - An open-source AI knowledge base Q&A system launched by FIT2CLOUD

What is MaxKB

MaxKB is an open-source AI knowledge base Q&A system launched by FIT2CLOUD, providing out-of-the-box functionality, supporting document upload, online document crawling, automatic text splitting, and vectorization. Users can quickly embed MaxKB into third-party business systems and enjoy the convenience brought by intelligent Q&A. It is model-neutral and compatible with a variety of large models, including local private and public large models, making it suitable for enterprises to build knowledge bases and improve user satisfaction.

Main features of MaxKB

• Out-of-the-box: Supports direct document upload and automatic crawling of online documents, enabling automatic text splitting and vectorization, and providing an intelligent Q&A interaction experience.
• Model-neutral: Can connect to a variety of large language models, whether local private models or domestic and international public models, such as Llama 3, Qwen 2, Tongyi Qianwen, OpenAI, etc.
• Flexible orchestration: Built-in powerful workflow engine, allowing users to orchestrate AI workflows to adapt to the needs of different business scenarios.
• Seamless embedding: Supports zero-code rapid embedding into third-party business systems, enabling existing systems to quickly gain intelligent Q&A capabilities.
• Multi-format support: Supports a variety of document formats, including TXT, Markdown, PDF, DOCX, HTML, etc.

Technical principles of MaxKB

• Large pretrained language model (LLM): MaxKB uses a large language model built with deep learning technology. The model has been trained on massive text data and has a high degree of language understanding and generation capability.
• Automated document processing: The system can automatically parse and process documents uploaded by users, including text segmentation and vectorization processing, so the model can index and retrieve information more efficiently.
• Retrieval-augmented generation (RAG) technology: MaxKB combines the advantages of retrieval systems and generative models, enhancing the generation capability of the language model by retrieving relevant information, thereby providing more accurate and richer answers.
• Model neutrality: By design, it supports integration with a variety of language models, including local private models and public models, ensuring the flexibility and extensibility of the system.
• Workflow engine: The built-in workflow engine allows users to customize and orchestrate AI workflows according to business needs, realizing automated and personalized intelligent Q&A services.
• Vector database application: Uses vector database technology, such as PostgreSQL/pgvector, optimizing document storage and retrieval efficiency and improving the system's response speed.
• Front-end and back-end separation architecture: The front end uses Vue.js to build a dynamic user interface, and the back end uses the Python/Django framework, ensuring the stability and maintainability of the system.
• LangChain framework integration: By integrating the LangChain framework, MaxKB can effectively manage and coordinate different AI models and services.

MaxKB project address

Application scenarios of MaxKB

• Enterprise knowledge base: Provides an internal knowledge Q&A system for enterprises, helping employees quickly find information and improve work efficiency.
• Customer service: As an intelligent assistant for customer support, it automatically answers common questions and reduces the workload of the customer service team.
• Education and training: Used for online education platforms, providing student Q&A services and assisting teachers with instruction.
• Product Documentation Support: Integrated into product documentation to provide users with instant technical support and answers to questions.
• Market Analysis: Analyze market data and trends to provide decision support for enterprises.