XNews

WeKnora: WeChat's enterprise LLM knowledge platform

WeKnora: WeChat's enterprise LLM knowledge platform

Tencent's WeChat team open-sourced WeKnora, an enterprise LLM knowledge platform that transforms scattered documents into RAG knowledge bases, ReAct agents, and auto-evolving wikis. It supports multi-tenant RBAC, private deployment, and integrates with various document sources.

Key insights

WeKnora converts scattered documents into three knowledge assets

WeKnora transforms raw documents into: RAG knowledge bases for Q&A with citation, ReAct agents for complex multi-step reasoning with tool use, and auto-evolving wikis that distill documents into linked Markdown knowledge bases with visual knowledge graphs. The wiki mode is a key differentiator, supporting up to 40,000 documents.

Enterprise features include multi-tenant RBAC and security

WeKnora supports multi-tenancy with four-level RBAC (Owner, Admin, Contributor, Viewer), tenant-level audit logs, API keys with scoped permissions, AES-256-GCM encryption for credentials, gRPC/Redis TLS, SSRF-hardened HTTP client, OIDC JWKS verification, and sandbox isolation. Knowledge bases can be edited with version diff and rollback.

Built on WeChat's production infrastructure

WeKnora is the core technology framework behind WeChat's dialogue open platform, with real production backing. It supports document sources from Feishu, GitLab, Notion, Yuque, DingTalk, WeCom, RSS, and more. It emphasizes private deployment with full data control.

Source posts