Flow Editor

Overview

The Flow Editor is a visual, no-code/low-code tool for building powerful data pipelines and automation workflows. With pre-built blocks and AI assistance, you can create complex data processing and AI workflows without writing code.

Getting Started

1. Create a New Flow

Go to the main page and start typing in the chat. If you don't know where to start, just greet your assistant with a "Hi". The chat will guide you through the process.

2. Add Data to Your Flow

Open the Node Library and click on the Select Data node. Navigate to the Data tab and click Add Data.

For detailed instructions on adding data, see the Data Module documentation.

3. Choose Your Workflow

Decide whether you want to train a model or run inference:

Train a model: Follow the Training guide
Run inference: Follow the Inference guide

4. Connect Your Nodes

To connect two nodes, drag from one node's dot to another node's green dot. A line will appear to show the connection.

Managing Connections:

Click on a line once to make it solid
Press DEL or click again to delete the connection

Keyboard Shortcuts:

CTRL+A: Select all nodes
DEL: Delete selected node or connection
Shift + Click: Multi-select or deselect from multi-select
Shift + Drag: Draw a box around nodes to multi-select

5. Configure Node Settings

Click on any node to select it. A navigation bar will appear on the right side where you can adjust the node's settings.

6. Run Your Flow

Click the RUN button (the green button at the bottom of the flow editor) to execute your workflow.

Important: While a flow is running, you cannot change node locations or settings. Wait until execution completes before making changes.

7. Deploy Your Flow

Deploy with one click to get your API endpoint. See the Deployment documentation for details.

Key Features

Pre-built Building Blocks

Access a comprehensive library of pre-defined building blocks that provide powerful algorithms. Navigate to the Node Library to see all available options.

Available Node Types

Select Data: Choose data from uploads or connectors in the Data Module
Apply Transformation: Clean, transform, and prepare your data
Train AI Model: Train custom machine learning models
Select Artifacts: Define what weights, label encoders, or other files AI models use during inference
API Input: Add validation schemas and test data to run inference on models or test deployed flows through APIs
Use AI Model: Run inference on AI models with public or custom trained models
Logic & Control Flow: Add conditional logic and loops to your workflows
Preview Output: Easily preview and visualize results of previous nodes

One-Click Deployment

Deploy your flows with a single click by navigating to the deployment tab. Scaling is handled automatically for each deployed flow.

AI Assistance

Use the AI Builder to assist you in creating your data and automation flows.

Common Use Cases

Data Pipelines: Automated ETL (Extract, Transform, Load) workflows
AI Workflows: End-to-end machine learning pipelines from data prep to deployment
API Services: Create custom APIs for data processing and AI inference
Automation: Automate repetitive tasks and business processes
Integration: Connect multiple systems and data sources

Tips for Building Flows

Start Simple: Begin with a basic flow and add complexity gradually
Use Preview Nodes: Add Preview Output nodes to debug your flow at each step
Test Before Deploy: Always run your flow with test data before deploying

Responsible Developers: ui:Sona, api:Sneha, Julia, Max, Usama.