Itzam SDK

Itzam provides tools for seamlessly integrating AI into your applications. With Itzam, you can manage AI workflows without mixing AI logic into your business code, allowing both technical and non-technical team members to control AI features through a simple dashboard interface.

Features

• 🤖 Model Hot Swap - Change AI models or settings instantly without code changes
• 💬 Prompt Hot Swap - Update prompts on the fly and compare outputs
• 💳 Unified Billing - Manage all AI spending in one place
• 🔌 Simple Integration - Add AI features with just a few lines of code
• 📎 Multimodal Support - Send files, images, and documents as attachments
• 🗃️ Context Management (coming soon) - Easily manage AI context including docs, images, and more
• 📐 Custom Rules (coming soon) - Create automated rules for model switching and workflow management

Installation

npm install itzam
# or
yarn add itzam
# or
pnpm add itzam

Quick Start

1. Create a workflow in the Itzam Dashboard
2. Get your API key and workflow slug
3. Integrate Itzam into your application:

import { Itzam } from "itzam";

// Initialize the client
const itzam = new Itzam("your-api-key");

// Or with custom base URL
const itzam = new Itzam("your-api-key", {
  basePath: "https://your-custom-api-url.com"
});

// Use your workflow
const response = await itzam.generateText({
  input: "Your input text here",
  workflowSlug: "your-workflow-slug",
});

console.log(response.output);

Working with Attachments

Send files, images, and documents with your requests:

// With base64 data
const response = await itzam.generateText({
  input: "What do you see in this image?",
  workflowSlug: "vision-workflow",
  attachments: [
    {
      file: "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8/5+hHgAHggJ/PchI7wAAAABJRU5ErkJggg==",
      mimeType: "image/png"
    }
  ]
});

// With URLs
const response2 = await itzam.generateText({
  input: "Analyze this document",
  workflowSlug: "document-analysis",
  attachments: [
    {
      file: "https://example.com/document.pdf",
      mimeType: "application/pdf"
    }
  ]
});

// With File objects (automatically converted to base64)
const fileInput = document.getElementById('fileInput') as HTMLInputElement;
const file = fileInput.files?.[0];

if (file) {
  const response3 = await itzam.generateText({
    input: "Process this file",
    workflowSlug: "file-processor",
    attachments: [
      {
        file: file, // SDK handles conversion automatically
        mimeType: file.type
      }
    ]
  });
}

Configuration

The Itzam SDK can be configured with the following options:

const itzam = new Itzam("your-api-key", {
  basePath: "https://your-custom-api-url.com" // Optional: Custom API base URL
});

Configuration Options

• basePath (optional): Custom base URL for the API. Defaults to the NEXT_PUBLIC_APP_URL environment variable if not provided.

Example: Building an AI Support Chat

Here's how to create an AI-powered support chat using Itzam:

1. Dashboard Setup

   • Create a "Support Chat" workflow in the Itzam Dashboard
   • Configure your preferred AI model (GPT, Claude, Gemini)
   • Set up your prompt and context
   • Get your workflow slug and API key

2. Code Integration

import { Itzam } from "itzam";

const itzam = new Itzam("your-api-key");

// Handle user messages
const handleUserMessage = async (userInput: string) => {
  const response = await itzam.generateText({
    input: userInput,
    workflowSlug: "support-chat",
  });

  return response.output;
};

Why Itzam?

• Separation of Concerns - Keep AI logic separate from your business logic
• No-Code Management - Empower non-technical team members to manage AI workflows
• Future-Proof - Easily switch between AI models as technology evolves
• Instant Updates - Change prompts and settings without redeploying your application

Documentation

For detailed documentation, visit our official documentation.

Support

• Email: [email protected]