6 min read

Getting Started with AIPI

This guide walks you through using Intuidy as the complete backend for your application — from setting up your account to deploying a working app with a fully managed BaaS behind it.


Prerequisites

Before you start, you'll need:

RequirementDescription
Vantage AccountAn active Intuidy Vantage account with API access enabled
Frontend ToolAny tool for building your UI — AI coding assistants (Cursor, Claude, ChatGPT), frameworks (React, Vue, Next.js), or even plain HTML/CSS/JS
HostingA place to deploy your frontend (Vercel, Netlify, AWS, Azure, Google Cloud, or your own servers)

No backend experience required. Intuidy is your backend. AIPI eliminates the need to build, host, or manage servers, databases, or infrastructure. You focus entirely on the application your users interact with.


Step 1: Set Up Your Vantage Account

  1. Sign up at the Vantage platform and complete the registration wizard
  2. Create a company — this is the data boundary for your application
  3. Configure integrations — connect the databases and services your app will use (see Connecting Integrations)

Configure Your Data Layer

Set up the database connections your application needs:

Tip: Every record stored through Vantage is structured, trackable, and immediately usable by AI agents. This means your application data can be analyzed, summarized, and acted on by AI from day one.


Step 2: Build Your Frontend

Use any technology to build your application's user interface:

ApproachToolsBest For
AI-assisted developmentCursor, Claude Code, ChatGPT Canvas, WindsurfFastest path — describe what you want and iterate
Framework-basedReact, Vue, Next.js, Svelte, AngularFull control over the application architecture
Low-codeAny drag-and-drop UI builderSimple applications or prototypes
NativeSwift, Kotlin, Flutter, React NativeMobile applications

Your frontend handles:


Step 3: Connect to AIPI

Point your frontend at the Vantage backend by configuring your AIPI connection:

  1. Get your API credentials from the Vantage Settings page
  2. Configure your application to send authenticated requests to the AIPI endpoints
  3. Test the connection by making a simple data query

Connection Pattern

  1. User interacts with your frontend UI
  2. App sends an AIPI request — authenticated, with relevant data
  3. Vantage processes the request — database query, workflow execution, AI analysis, or any combination
  4. Structured response returned to your app, ready to display

Step 4: Build Your Workflows

Create the backend logic your application needs. Because AIPI is fully headless, you have two paths:

Option A: Let Your Agentic IDE Build Them

Your AI coding assistant (Cursor, Claude, Windsurf, etc.) can create and manage workflows programmatically through the REST API:

bash
curl -X POST https://your-instance.com/api/v1/workflows \
  -H "Authorization: Bearer vntg_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Customer Lookup",
    "nodes": [
      { "node_type": "triggers/workflowInput", "node_role": "getter", "label": "Input" },
      { "node_type": "database/query", "node_role": "setter", "label": "DB Lookup" },
      { "node_type": "termination/workflowOutput", "node_role": "terminator", "label": "Output" }
    ],
    "edges": [
      { "source_node_index": 0, "target_node_index": 1 },
      { "source_node_index": 1, "target_node_index": 2 }
    ]
  }'

Your IDE can discover available node types via GET /api/v1/node-types, then compose and deploy workflows entirely from code — no browser required.

Option B: Build Manually in the Visual Editor

  1. Navigate to Workflows in Vantage
  2. Create a new workflow for each backend operation
  3. Use Workflow Input nodes to receive data from your AIPI calls
  4. Add processing nodes (database queries, AI analysis, transformations)
  5. Use Workflow Output nodes to return results to your application

Example: Customer Lookup Workflow

StepNodePurpose
1Workflow InputReceives the customer ID from your AIPI call
2Database QueryLooks up the customer record
3AI EnrichmentGenerates a customer summary using an AI model
4Workflow OutputReturns the customer data and AI summary to your app

See Building Workflows for the full workflow guide.


Step 5: Deploy Your Application

Host your frontend wherever you prefer — AIPI works with any hosting provider:

ProviderHow to Deploy
VercelConnect your Git repository for automatic deployments
NetlifyDrag-and-drop your build folder or connect Git
AWSUse S3 + CloudFront, Amplify, or EC2
AzureAzure Static Web Apps or App Service
Google CloudFirebase Hosting or Cloud Run
Self-hostedAny web server (Nginx, Apache, etc.)

AIPI is hosting-agnostic. Your frontend sends requests to the Vantage API regardless of where it's deployed. No vendor lock-in.


Step 6: Configure AI Agents (Optional)

Deploy AI agents that can act on your application's data:

  1. Navigate to Agent Hub in Vantage
  2. Create a new agent with the Agent Builder
  3. Assign skills, connect data resources, and write a system prompt
  4. Deploy the agent — it becomes accessible through AIPI

Your application can send messages to the agent, and the agent can:

See Agent Builder for the full agent setup guide.


What Happens Next

Once your application is connected to AIPI, you can incrementally add capabilities:

Want ToHow
Add a new data sourceConnect another integration in Vantage — it becomes accessible through AIPI immediately
Automate a processBuild a workflow in the Workflow Editor and trigger it from your app
Add AI featuresDeploy an agent or use AI workflow nodes to analyze your data
Build custom integrationsUse the API Builder to connect any REST API
Scale upAIPI scales with the Vantage platform — no infrastructure changes needed

Next Steps