@regia-ai/js-sdk

@regia-ai/sdk

JavaScript/TypeScript SDK to interact with the Regia AI services and APIs.

🚀 Installation

npm install @regia-ai/js-sdk

---

📦 Basic Usage

TypeScript Example

import { Vision } from '@regia-ai/js-sdk';
import { z } from 'zod';

// Initialize the Vision client with your token
const vision = new Vision("your_token_here");


// Define your schema using Zod
const schema = z.object({
  usageKwh: z.number().describe("The total usage in kWh"),
  dueDate: z.string().describe("The due date of the invoice in the format YYYY-MM-DD"),
  totalAmount: z.number().describe("The total amount of the invoice"),
})

// Extract data from the PDF file
const { usageKwh, dueDate, totalAmount } = await vision.extract('./sample.pdf', { schema });

// Log the extracted data
console.log("Usage kWh: ", usageKwh);
console.log(`Due date: ${dueDate}`);
console.log(`Total amount: ${totalAmount}`);

---

🧠 Schema Definition

You define the structure of the extracted data using Zod, a TypeScript-first schema declaration and validation library.

This allows the server to understand what fields to extract using generative AI.

You can also pass a JSON Schema string to the schema parameter.

---

🔐 Authentication

All requests require a bearer token passed during client initialization:

const client = new Vision('your-api-token')

---

📘 API Reference

For more information checkout the Developer Documentation.

Vision(token, options)

Option	Type	Required	Description
token	string	✅	Your API token

---

client.extract(file, { schema })

Extracts structured data from a media file.

Parameter	Type	Required	Description
file	string	✅	Path to the local file or a base64-encoded string.
schema	ZodSchema	✅	The schema defining the desired fields to extract.

---

✅ Requirements

• Node.js >= 18
• Zod for schema definition
• Access to Regia's PDF Extractor API

---