Your First Workflow

Create your first AI workflow step-by-step

This guide walks you through creating your very first AI workflow in BroxiAI, from start to finish.

What We'll Build

You'll create a simple but powerful workflow that:

  • Takes user input

  • Processes it with AI

  • Returns an intelligent response

  • Demonstrates core BroxiAI concepts

Before We Start

Make sure you have:

  • ✅ A BroxiAI account

  • ✅ An OpenAI API key (or another AI provider key)

  • ✅ Completed the Installation & Setup

Step 1: Access the Workspace

  1. Log into BroxiAI

  2. Open the Workspace

    • Click on "New Flow" from your dashboard

    • You'll see the visual workflow builder

  3. Familiarize Yourself

    • Left Sidebar: Component library

    • Center: Workspace canvas

    • Right Panel: Component properties and playground

Step 2: Plan Your Workflow

Understanding the Flow

Every BroxiAI workflow follows this pattern:

Input → Processing → Output

For our first workflow:

  • Input: User message

  • Processing: AI model generates response

  • Output: Display the response

Choose Your Components

We'll use three essential components:

  1. Chat Input: Receives user messages

  2. OpenAI Model: Processes with AI

  3. Chat Output: Displays the response

Step 3: Add Components

Add Chat Input

  1. Find the Component

    • Look in the left sidebar under "Input & Output"

    • Drag "Chat Input" onto the canvas

  2. Position the Component

    • Place it on the left side of your canvas

    • This will be your workflow's starting point

  3. Configure Basic Settings

    • Click on the component to open properties

    • Set a descriptive name: "User Message"

Add OpenAI Model

  1. Locate AI Models

    • Find "AI Models" section in the sidebar

    • Drag "OpenAI" component to the center

  2. Configure the Model

    • Model: Select "gpt-3.5-turbo" (cost-effective)

    • Temperature: Set to 0.7 (balanced creativity)

    • Max Tokens: Set to 500 (reasonable response length)

  3. Add System Prompt

    You are a helpful AI assistant. Provide clear, accurate, and friendly responses to user questions. Keep responses concise but informative.

Add Chat Output

  1. Find Output Component

    • In "Input & Output" section

    • Drag "Chat Output" to the right side

  2. Configure Display

    • Name it "AI Response"

    • Enable "Show in Playground" for testing

Step 4: Connect Components

Understanding Connections

  • Output handles (⚪): Data flows out

  • Input handles (⚪): Data flows in

  • Drag and drop to create connections

Make the Connections

  1. Connect Input to AI Model

    • Click and drag from Chat Input's output handle

    • Connect to OpenAI Model's "input" port

    • You'll see a line connecting them

  2. Connect AI Model to Output

    • Drag from OpenAI Model's output handle

    • Connect to Chat Output's input port

Verify Your Flow

Your canvas should show:

Step 5: Test Your Workflow

Open the Playground

  1. Click Playground Button

    • Located in the top-right corner

    • The chat interface will appear

  2. Start Your First Conversation

    • Type: "Hello! Can you introduce yourself?"

    • Press Enter or click Send

    • Watch the AI respond!

Try More Examples

Test with different types of inputs:

Simple Questions

What's the capital of France?
How does photosynthesis work?
What's 25 × 17?

Creative Requests

Write a short poem about coding
Tell me a fun fact about space
Create a recipe for chocolate cake

Problem Solving

How do I stay motivated while learning?
What's the best way to organize my time?
Explain machine learning in simple terms

Step 6: Understand What Happened

Data Flow

  1. User Input: Your message enters through Chat Input

  2. AI Processing: OpenAI model generates a response

  3. Output Display: Response appears in Chat Output

Behind the Scenes

  • BroxiAI handles API calls to OpenAI

  • Manages authentication and error handling

  • Processes and formats the response

  • Maintains conversation context

Step 7: Enhance Your Workflow

Add Conversation Memory

  1. Add Memory Component

    • Find "Conversation Buffer Memory" in Components

    • Place it between Chat Input and OpenAI Model

  2. Configure Memory

    {
      "memory_key": "chat_history",
      "max_token_limit": 2000,
      "return_messages": true
    }
  3. Update Connections

    • Disconnect direct connection

    • Connect: Input → Memory → AI Model → Output

Test Memory

Try this conversation:

You: "My name is Alex"
AI: "Hello Alex! Nice to meet you."
You: "What's my name?"
AI: "Your name is Alex."

Step 8: Save and Organize

Save Your Work

  1. Name Your Flow

    • Click on "Untitled Flow" at the top

    • Rename to "My First Workflow"

  2. Add Description

    • Click the settings icon

    • Add: "My first BroxiAI workflow - a simple AI chat assistant"

  3. Save Changes

    • Changes are automatically saved

    • Look for the "Saved" indicator

Organize in Folders

  1. Create a Folder

    • Go to "My Projects"

    • Click "New Folder"

    • Name it "Learning Projects"

  2. Move Your Workflow

    • Drag your workflow into the folder

    • Keep your workspace organized

Understanding Key Concepts

Components

  • Building blocks of your workflow

  • Each has specific functions and purposes

  • Can be configured with parameters

  • Connect to form data processing pipelines

Data Types

  • Text: Messages, responses, prompts

  • Data: Structured information

  • Files: Documents, images, audio

  • Messages: Chat conversation format

Execution Flow

  • Workflows execute when triggered

  • Data flows through connected components

  • Each component processes and transforms data

  • Results appear in output components

Common First-Timer Issues

Component Won't Connect

Problem: Red error when trying to connect Solution: Check data type compatibility

AI Not Responding

Problem: No response from AI model Solution: Verify API key in Settings → API Keys

Empty Response

Problem: AI returns empty or error message Solution: Check system prompt and input format

Slow Response

Problem: Takes too long to respond Solution: Try a faster model like gpt-3.5-turbo

Next Steps

Congratulations! You've created your first BroxiAI workflow. Here's what to explore next:

Immediate Next Steps

  1. Experiment with Settings

    • Try different temperature values

    • Adjust max tokens

    • Modify the system prompt

  2. Test Edge Cases

    • Very long messages

    • Empty inputs

    • Special characters

Learning Path

  1. Understanding Components

    • Learn about different component types

    • Understand when to use each one

  2. Workspace Tour

    • Master the interface

    • Learn productivity tips

  3. Configuration Guide

    • Advanced settings

    • Environment management

Build More Workflows

Resources

  • Template Library: Pre-built workflows to learn from

  • Community Examples: See what others have built

  • Help Center: Get answers to common questions

Troubleshooting

Having issues? Check our Troubleshooting Guide or:

  • Community Forums: Ask questions and get help

  • Discord Channel: Real-time community support

  • Support Team: Direct assistance when needed


Last updated