Quickstart
Go from zero to a fully functional backend in under five minutes. This guide walks you through setting up your account and connecting your IDE to start building with AI.
Prerequisites
You need a DYPAI account to continue. If you don't have one yet, sign up for free at dypai.ai.
1. Set up your Organization
Organizations are the top-level container for your projects and team. If you just signed up, you'll need to create one.
Sign in
Go to dypai.ai and sign in to your dashboard.
Create Organization
If you don't have an organization yet, click New Organization. Give it a name (e.g., your company or project name).
2. Create a Project
Each project is an isolated environment with its own database, authentication, and storage.
New project
From your organization dashboard, click New Project.
Provisioning
Choose a name and plan. DYPAI will automatically set up the initial version of your project, including the database, infrastructure, and core services. This takes just a few seconds.
3. Connect via MCP (The AI Path)
Instead of manually creating your endpoints, database tables, and logic, the fastest way to build with DYPAI is by connecting your project to an AI-powered IDE (like Cursor or Claude Desktop) using our Model Context Protocol (MCP).
Find the MCP Button
In your project dashboard, look for the MCP button in the top header. It's always available so you can quickly access your connection details.
Copy your Configuration
Click the button to open the connection dialog. You can copy your unique MCP URL or the complete JSON configuration snippet for your IDE.
Configure your IDE
Paste the configuration into your IDE settings (e.g., .cursorrules or MCP settings in Claude). This allows your AI assistant to "see" your project and execute commands.
Start Building
Now you can simply ask the AI: "Create a 'users' table with email and name columns" or "Set up a WhatsApp integration workflow".
4. Install the SDK (recommended)
Once MCP is connected, install the SDK so your frontend can talk to your DYPAI backend.
Install the DYPAI package
In your frontend app root, install the official SDK:
npm install @dypai-ai/client-sdk
Verify your setup
You can now combine MCP prompts (build backend) with frontend calls through the SDK.
Backend-only with MCP?
If you are not connecting a frontend yet, you can temporarily skip SDK installation and use MCP only.
5. Configure .env variables in your app
Use the same structure that appears in the Connection modal of your DYPAI project.
Choose the mode you need (jwt for user-facing app flows, api_key for server-side flows) and copy the snippet for your framework:
# .env.local
NEXT_PUBLIC_DYPAI_URL=https://<project-id>.dypai.dev
DYPAI_API_KEY=anon-xxxxxxxxxxxxxxxx
# Keep the API key only in Next.js server-side codeFor browser/user auth flows, initialize the SDK with only your project URL:
import { createClient } from '@dypai-ai/client-sdk'
export const dypai = createClient(process.env.NEXT_PUBLIC_DYPAI_URL!)
Security
Keep api_key in server-only env vars. Do not expose project API keys in browser code or NEXT_PUBLIC_* variables.
Not sure which auth flow fits your app?
Read Auth Flows before wiring signup, invitations, or backend-only calls.
Next steps
Now that you are connected, you can explore the core features of the platform:
- Using the Client SDK — Connect your frontend to your new backend.
- Database Management — View and edit your data visually.
- API Builder — Create custom logic and workflows.
- Authentication — Secure your app and manage users.