Enterprise multi-tenant AI knowledge base with RAG pipeline. Hybrid retrieval (pgvector + Elasticsearch BM25) + RRF + Rerank, MCP Server, Canal binlog sync, RabbitMQ async processing, Redis cache, Spring Boot 3.5 + Java 17
View on GitHub →
Try with Claude — $10 free →
Quick Install
Copy the config for your editor. Some servers may need additional setup — check the README.
Add to claude_desktop_config.json:
{
"mcpServers": {
"knowhub": {
"command": "npx",
"args": [
"-y",
"csicetus/knowhub"
]
}
}
}
Run in terminal:
claude mcp add knowhub npx -y csicetus/knowhub
Add to .cursor/mcp.json:
{
"mcpServers": {
"knowhub": {
"command": "npx",
"args": [
"-y",
"csicetus/knowhub"
]
}
}
}
README Excerpt
KnowHub 是一个企业级多租户 AI 知识库平台，支持文档上传、智能向量化存储和 RAG 智能问答。 用户可上传 txt/pdf/md 文档，系统自动解析分块、生成 Embedding 存入向量数据库，生成 Embedding 存入向量数据库，通过混合检索（pgvector 向量检索 + ES 全文检索 BM25）经 RRF 融合和 Rerank 精排，获得基于文档内容的精确回答，支持 SSE 流式输出打字机效果。 | 层次 | 技术选型 | 选型理由 | |-------|---------------------------|----------------------------------------|