WeKnora
Turn public and internal documents into searchable RAG, reasoning agents, and automated wikis.
Unify city documents, policies, and municipal records into queryable RAG networks and automated agents.
WeKnora is a Go-based open-source framework that converts unstructured enterprise documents into searchable RAG bases, interactive agents, and connected wikis. It enables public sector and urban management teams to query policy archives, automate multi-step document tasks, and maintain structured knowledge networks. You can deploy it completely within your private network or on cloud infrastructure.
What it is
The platform parses over ten document formats—including Office files, PDFs, and mind maps—and syncs automatically with tools like Confluence, GitLab, and Feishu. It runs reasoning agents inside isolated Docker or E2B sandboxes with desktop and terminal access to process complex workflows. You can also hook into 29 model providers or local Ollama instances while tracking token usage via built-in Langfuse tracing.
Who it's for
It suits municipal departments, policy analysts, and IT administrators needing strict data sovereignty and automated knowledge retrieval across scattered file repositories. It is not designed for non-technical users seeking a simple desktop note-taking application without server configuration. Organizations without dedicated IT personnel to manage Docker containers or vector databases should avoid self-hosting it.
Key features
Automated Sandbox Execution
Agents execute complex multi-step code and browser tasks inside persistent Docker or E2B sandboxes with desktop and terminal access.
Multi-Source Document Sync
Automatically ingest and update knowledge bases from platforms like Confluence, Notion, GitLab, and enterprise drives.
Integrated Knowledge Graphing
Link document chunks into visual wiki graphs using Neo4j backends for clear structural relationships.
Multi-Workspace Access Control
Manage public sector team permissions across four RBAC roles with per-resource ownership and audit logs.
Up and running in 3 steps
Our honest verdict
What's great
- Connects directly to 29 AI model vendors alongside local Ollama instances for offline deployments.
- Parses Office files in-process without requiring external file-conversion microservices.
- Runs automated background shell jobs and agent workflows with explicit approval checkpoints.
- Integrates directly with enterprise channels like WeCom, Slack, Telegram, and custom web widgets.
Where it falls short
- Requires significant server infrastructure and technical skill to manage full vector, graph, and tracing backends.
- No official pre-built desktop installers are available, requiring manual compilation from source for desktop use.
- Must be isolated behind internal private networks and firewalls due to direct public deployment security risks.
- No explicit license is stated in the repository metadata, creating legal ambiguity for enterprise deployment.
Pricing & license
WeKnora is free to download from GitHub, though no explicit license file is included in the repository. You can run it on your own hardware or deploy it via cloud templates like Tencent Cloud Lighthouse. Your main expense will be hosting infrastructure, vector storage, and model API tokens.
Open remote AI-training roles
767 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
Use it, or have us set it up
FAQ
Can I run WeKnora completely offline for government compliance?
Yes, you can run WeKnora on private infrastructure using local Ollama models and self-hosted vector databases.
What document formats does WeKnora support?
It parses over 10 formats including PDF, Word, Excel, images, and XMind files directly in-process.
How does WeKnora handle secure multi-department data?
It provides multi-workspace RBAC with four predefined roles, resource ownership, and full audit logs.
