Intro to Qelos
What is Qelos?
Qelos is an AI-first application layer gateway. It is the single backend layer that any application plugs into to get a fully managed REST API, AI agents, authentication, multi-tenant workspaces, permissions, and an events log — all accessible through a unified SDK and per-framework integrator middleware.
Your app stays in your own framework. You define your data model with blueprints in the admin panel, and the gateway generates CRUD endpoints, permission enforcement, and audit logging for you.
The three roles Qelos plays
🛠️ Admin panel for applications
A dashboard where developers and operators manage their app's data model, AI agents, users, workspaces, permissions, and events.
🌐 API gateway for apps
A RESTful API backed by MongoDB that apps consume via the SDK. Blueprints define the data model; the gateway exposes CRUD, AI chat, auth, and more.
🧩 A plugin for every framework
Integrator middleware (@qelos/integrator-*) makes the Qelos API the API for your app with a single import.
What Qelos brings to your app
🧱 Blueprints → REST API
Define your data model in the admin panel; get full CRUD endpoints, validation, and permission checks generated automatically.
🤖 AI agents & chatbots
Create, configure, and embed AI agents. Switch providers (OpenAI, Anthropic, Gemini) without changing your product architecture.
🔐 Auth, permissions & workspaces
Multi-tenant workspaces, social login, cookie/token lifecycle, roles, and resource-level permissions enforced at the gateway.
🧰 Unified SDK + CLI
One TypeScript or Python SDK for every capability, plus @qelos/cli for scaffolding, agents, and IDE rules generation.
🔌 Per-framework middleware
Drop-in integrators for Next.js, Nuxt, Express, Fastify, NestJS, and FastAPI — one import wires up auth, workspace, and SDK on every request.
📜 Events & audit log
Every API call, auth event, and AI interaction is recorded. Forward to webhooks or query through the SDK.
How developers use Qelos
- Install the CLI:
npm install -g @qelos/cli - Initialize in their project:
qelos init(detects framework, scaffolds config) - Install the integrator:
npm install @qelos/integrator-next(or express, nuxt, fastify, nest, fastapi) - Use the SDK in their code:
sdk.entities('products').find(),sdk.ai.agents.chat(), etc. - Manage everything from the admin panel or the CLI.
import QelosSDK from '@qelos/sdk';
const sdk = new QelosSDK({
appUrl: process.env.QELOS_APP_URL!,
apiToken: process.env.QELOS_API_TOKEN!,
});
const products = await sdk.entities('products').find({ status: 'active' });
const reply = await sdk.ai.agents.chat('agent-id', 'How can I help?');How to get started
Pick the path that matches what you're trying to do:
- Getting Started as an Integrator — plug an existing app into a Qelos instance using the CLI, the integrator for your framework, and the SDK.
- Installation — clone the repo and run Qelos itself locally (for contributors and self-hosting).
- Create Blueprints — model your application's data and get CRUD endpoints automatically.
- Configure Your First AI Agent — wire up an agent end-to-end and call it from your app.
- Deployment — ship Qelos to production with Docker Compose or Kubernetes / Helm.
HTTP API and the gateway
Every product app talks to Qelos through a single gateway: browser traffic and API calls hit one origin; paths such as /api/me, /api/blueprints/.../entities, /api/ai/..., and /api/events are proxied to internal services (auth, no-code, AI, plugins, and others). You do not call those microservices directly from client code.
For a concise map of those routes, request headers, example curl commands, and @qelos/sdk equivalents organized by domain, see the Gateway endpoint reference. The full topic pages under API Reference expand request and response shapes for each area.