TensorifyDocs

Quick Start Guide

Get your first Tensorify workflow running in under 5 minutes.

Prerequisites

Before you begin, make sure you have:

  • A Tensorify account (sign up here)
  • A modern web browser (Chrome, Firefox, Safari, or Edge)

Step 1: Create a New Workflow

  1. Log into your Tensorify dashboard
  2. Click "New Workflow" in the sidebar
  3. Give your workflow a name (e.g., "My First Workflow")
  4. Click "Create"

You'll be taken to the visual canvas editor.

Step 2: Add Your First Plugin

The plugin sidebar on the left shows all available plugins. Let's start with a simple example:

  1. Search for "OpenAI" in the plugin sidebar
  2. Drag the OpenAI plugin onto the canvas
  3. Click on the node to configure it

Configure the Plugin

In the settings panel:

  • Model: Select gpt-4 or gpt-3.5-turbo
  • System Prompt: Enter "You are a helpful assistant"
  • API Key: Your OpenAI API key (stored securely)

Step 3: Add an Output

  1. Search for "File Writer" in the plugins
  2. Drag it onto the canvas, to the right of the OpenAI node
  3. Connect them by dragging from the OpenAI output handle to the File Writer input

Connect the Nodes

Nodes are connected through handles:

  • Output handles (right side) emit data
  • Input handles (left side) receive data

Click and drag from an output to an input to create a connection.

Step 4: Export Your Code

Once your workflow is connected:

  1. Click the "Export" button in the toolbar
  2. Choose "Download Python"
  3. You'll receive a zip with:
    • main.py - The executable script
    • utils.py - Helper functions
    • requirements.txt - Python dependencies

Step 5: Run Locally

# Extract the downloaded zip
unzip my-workflow.zip
cd my-workflow

# Install dependencies
pip install -r requirements.txt

# Set your API key
export OPENAI_API_KEY="your-key-here"

# Run the workflow
python main.py
✅

Congratulations! You've just built and run your first Tensorify workflow.

Alternative: Cloud Execution

Instead of running locally, you can execute directly on Tensorify's cloud:

  1. Click "Execute" instead of "Export"
  2. Choose your execution environment (CPU or GPU)
  3. Click "Run"
  4. View logs in real-time

Cloud execution is perfect for:

  • GPU-intensive ML workloads
  • Scheduled/automated runs
  • Team collaboration

Next Steps

Now that you've built your first workflow, explore:

Need help? Join our Discord community or email [email protected]