At a Glance:
NocoBase is an open-source AI + no-code platform for building business systems, featuring a WYSIWYG no-code interface, AI coding agent compatibility, and built-in AI employees that operate inside business workflows with field-level permission controls.
Overview:
NocoBase is an open-source AI + no-code platform designed to accelerate the construction of business systems. It combines a production-proven infrastructure with a visual WYSIWYG interface, allowing both AI coding agents and people to build and manage applications collaboratively. The platform includes built-in modules for data models, permissions, workflows, and audit logs, while also integrating AI employees that can execute tasks and make decisions within business processes. NocoBase uses a data-model driven architecture that decouples business data from the UI and supports external databases and third-party APIs as data sources.
Key Decision Points:
AI and no-code collaboration: AI coding agents work through a provided CLI and skills, while people configure visually via a WYSIWYG interface, allowing both to modify the same system.
Agent ecosystem integration: External agents and platforms like n8n, Dify, Coze, and messengers including Slack, Telegram, and WhatsApp can connect through MCP, HTTP APIs, and the CLI with defined permission boundaries.
AI employees with permissions: AI employees get dedicated roles with field-level read and write permissions in the standard access control system, and their actions are recorded in audit logs.
Data ownership and sources: Business data stays in the user's own database with support for the main database, external databases, and third-party APIs as data sources, avoiding platform lock-in.
Plugin architecture: The microkernel design treats every feature as a composable plugin, allowing custom and official plugins to be mixed without losing architectural control.
Core Features:
WYSIWYG no-code interface: Users can switch between usage and configuration modes to visually configure data models, pages, workflows, and permissions.
AI agent development: Mainstream coding agents, such as Claude Code, Cursor, and Codex, can handle setup, development, migration, and release through a full CLI.
AI employees: AI can perform tasks on the front-end, such as analysis and form filling, and on the back-end, such as document recognition and automatic task routing, while integrated into business workflows.
Data-model driven architecture: Relational data structures are kept separate from the UI, with the ability to use multiple data sources including external databases and third-party APIs.
Composable plugin system: A microkernel architecture where all functionality, including features built by AI or manually, is added through plugins with shared conventions.
Audit logs: Every data change and workflow trigger is logged, providing traceability for both AI and human actions.
Use Cases:
Rapid business system development: Developers and non-developers can collaboratively build working systems where coding agents generate initial configurations and users refine them visually.
Workflow automation with AI: Organizations can embed AI employees into business processes to handle tasks like risk monitoring, document recognition, and automated decision-making based on business context.
Multi-source data applications: Users can create applications that combine data from a primary database, external databases, and third-party APIs within a unified, visually configurable data model.
Open-Source Alternative Value:
As an open-source platform, NocoBase allows teams to build on production-proven, transparent infrastructure without regenerating core components as black-box code. Its architecture supports using your own database for all business data and avoids platform lock-in. The plugin system enables sustainable customization through composable modules, and the standard CLI and API interfaces allow integration with the broader agent ecosystem while keeping AI actions governed by the same permission system as human users. The project's AI employee concept provides a directly integrated, permission-controlled approach to embedding AI into business workflows.
