zgiai/zgi

ZGI is an open-source platform for building AI applications. Its intuitive interface combines workflow design, agent orchestration, dataset management, and model integration—allowing you to quickly move from prototype to production.

What it solves

ZGI provides a unified workspace for teams to move beyond simple chat interfaces and build AI agents that perform actual work. It solves the fragmentation of agent development by integrating configuration, visual orchestration, data retrieval, and execution environments into a single self-hostable platform.

How it works

ZGI operates as an Agent Runtime platform where users build agents and executable workflows on a visual canvas. These agents are connected to models, knowledge bases, databases, and reusable skills. Once configured, they can be executed via a WebApp, an internal app center, or an API. The system includes a model gateway for routing and quota management, and a sandboxed runtime to execute tools and code safely.

Who it’s for

It is designed for teams and organizations that need to build, publish, and operate governable AI agents and complex executable workflows within their own infrastructure.

Highlights

  • Visual Workflow Orchestration: Create complex logic including loops, branches, approvals, and HTTP requests on a visual canvas.
  • Advanced Retrieval: Supports semantic, full-text, hybrid, and knowledge-graph retrieval with reranking.
  • Sandboxed Execution: Runs tools and code in an isolated environment for security.
  • Model Gateway: Centralized management of providers, credentials, routing policies, and pricing.
  • Self-Hosted Runtime: Full stack (API, sandbox, runner, PostgreSQL, Redis) can be run locally or on private infrastructure.

Related

  • Dispatch
    Runtime: Scaling Agentic Workflows Across the Enterprise

    An exploration of Runtime, a sandboxed infrastructure for deploying coding agents across non-engineering teams while maintaining strict security and governance.

  • Project
    opensandbox-group/OpenSandbox

    A general-purpose sandbox platform for AI applications that provides secure, isolated environments for code execution, agent evaluation, and RL training.

  • Project
    madarco/agentbox

    A tool for running multiple AI agents in parallel within sandboxed VMs or Docker containers, either locally or in the cloud, with persistent IDE and shell access.

  • Project
    erthorpabar/pabar_agent_runtime

    An AI agent runtime framework that integrates chat, RAG, and agentic loops to enable professional automation tasks across desktop and mobile platforms.

  • Project
    xerrors/Yuxi

    A multi-tenant knowledge agent platform for private deployment that integrates RAG, knowledge graphs, and multi-agent orchestration to turn AI conversations into task deliverables.