🤖 AI Workflow Studio for CrewAI

A professional, no-code interface for building and executing multi-agent AI workflows with CrewAI.

Create complex agent teams, define collaborative tasks, and orchestrate intelligent workflows—all through an intuitive visual builder. Perfect for rapid prototyping, automation, and production AI applications.

---

🎯 Features

Core Capabilities

• Visual Agent Builder - Create agents with custom roles, goals, and personas
• Pre-built Templates - Quick-start with Content Team, Data Analysis, Research workflows
• Task Orchestration - Define sequential or parallel tasks with agent assignments
• Live Code Generation - Auto-generate production-ready Python code
• Real-time Execution - Run workflows directly from the interface with live feedback
• Multi-LLM Support - Switch between Groq models (Mixtral, Llama2, Gemma)

Professional Features

• Agent Persistence - Save and reuse custom agent configurations
• Template System - Create workflow templates for your team
• Code Export - Download generated code for deployment
• Session Management - Maintain workflow state across interactions
• Error Handling - Graceful error messages and validation
• Verbose Logging - Monitor agent thinking process in real-time

---

🚀 Quick Start

Prerequisites

• Python 3.8 or higher
• Groq API key (get one free)

Installation

# Clone the repository
git clone https://github.com/guiedup/Streamlit-Crew-AI.git
cd Streamlit-Crew-AI

# Install dependencies
pip install -r requirements.txt

# Run the application
streamlit run app.py

The application will open at http://localhost:8501

First Steps

1. Get API Key: Visit console.groq.com/keys
2. Enter API Key: Paste it in the Settings panel (⚙️) on the left
3. Load Template: Select a template (Content Team, Data Team, etc.) to get started
4. Build Your Workflow: Add agents, define tasks, and run!

---

📖 User Guide

Creating a Workflow

Step 1: Add Agents

Two ways to populate your workflow:

Using Pre-built Agents

• Click "Premade Agents" to expand
• Select agents like "🔍 Researcher", "📊 Analyst"
• Agents appear in your workflow on the right

Creating Custom Agents

• Expand "New Agent" in the top panel
• Fill in:
  • Emoji: Visual identifier for your agent
  • Role: Agent's job title (e.g., "Senior Developer")
  • Goal: What the agent aims to accomplish
  • Backstory: Agent's expertise and experience (context for LLM)
• Click "➕ Save Agent"
• Agent is now available in "My Agents"

Step 2: Configure Tasks

For each agent in your workflow:

• Expand the agent's task configuration
• Write Task Description: What needs to be done
• Specify Expected Output: What success looks like
• Tasks execute in order top-to-bottom

Step 3: Execute or Export

Option A: Run Live

• Click "⚡ Execute" tab
• Click "🚀 Execute Pipeline"
• Watch agents work in real-time
• View results in the output panel

Option B: Export Code

• Click "📄 Generate the Code" tab
• Click "🔄 Generate Python Code"
• Copy the generated code
• Use in your own Python scripts

Using Templates

Pre-configured workflows for common scenarios:

| Template | Agents | Use Case | |----------|--------|----------| | Content Team | Writer + SEO Specialist | Blog posts, articles with SEO optimization | | Data Team | Data Analyst + Data Scientist | Analytics, predictions, data pipelines |

To load a template:

1. Select from "🏆 Crew Templates" dropdown in the sidebar
2. Click "🔄 Load Template"
3. Templates pre-populate agents and tasks
4. Customize as needed

---

⚙️ Configuration

LLM Settings (Sidebar)

• Model: Select from available Groq models
  • mixtral-8x7b-32768 (best performance)
  • llama2-70b-4096 (balanced)
  • gemma-7b-it (fastest)
• Temperature: 0.0-1.0 (lower = more deterministic)
• Max Tokens: Maximum output length per task

API Configuration

# .streamlit/secrets.toml (optional)
[groq]
api_key = "your-groq-api-key-here"

---

🏗️ Architecture

Project Structure

Streamlit-Crew-AI/
├── app.py                 # Main Streamlit application
├── requirements.txt       # Python dependencies
├── tests.py               # Test suite
├── .gitignore            # Git ignore rules
├── LICENSE               # MIT License
└── README.md             # This file

Key Components

app.py (277 lines)

• Streamlit UI orchestration
• Agent and task management
• Code generation engine
• Workflow execution handler
• Session state management

Agent Management

• Pre-defined agents (Research, Analysis templates)
• Custom agent creation and persistence
• Agent configuration UI

Task Orchestration

• Sequential task assignment to agents
• Task description and output specification
• Workflow state tracking

Code Generation

• Auto-generates CrewAI Python code
• Exports ready-to-use agent definitions
• Copyable for external execution

---

💡 Use Cases

1. Content Creation Pipeline

Goal: Blog articles with SEO optimization

Writer Agent → (Researches topic, creates outline)
    ↓
SEO Agent → (Optimizes for keywords, improves readability)
    ↓
Output: SEO-optimized blog post

2. Data Analysis Workflow

Goal: Extract insights from datasets

Data Analyst → (Loads data, performs EDA)
    ↓
Data Scientist → (Builds predictive model, generates insights)
    ↓
Output: Analysis report with predictions

3. Research & Summaries

Goal: Comprehensive research on a topic

Research Agent → (Finds sources, synthesizes info)
    ↓
Verification Agent → (Fact-checks, cites sources)
    ↓
Output: Verified research summary

4. Customer Support

Goal: Multi-agent support system

Support Agent → (Handles common questions)
    ↓
Escalation Agent → (Routes complex issues)
    ↓
Output: Resolved ticket or escalation log

---

🔐 Security

API Key Handling

• Never commit API keys to git
• Use Streamlit's secrets.toml for local development
• Use environment variables in production
• API key input field is password-masked

Best Practices

• Rotate API keys regularly
• Use minimal-permission service accounts
• Audit usage on Groq console
• Monitor token consumption

---

🧪 Testing

Run the included test suite:

python tests.py

Test coverage includes:

• Agent creation and validation
• Task configuration
• Workflow state management
• Code generation logic

---

🐛 Troubleshooting

"API Key Needed!"

• Problem: Button won't execute without API key
• Solution: Paste valid Groq API key in ⚙️ Settings panel
• Get API key: https://console.groq.com/keys

"Empty Workflow"

• Problem: Agents not appearing
• Solution: Add agents from "Premade Agents" or create custom agents
• Tips: Try loading a template first

Agent Execution Errors

• Problem: Workflow execution fails
• Likely causes:
  • API key quota exceeded
  • Task description too vague (LLM can't understand)
  • Network connectivity issues
• Solution:
  • Check API key is valid
  • Review Groq console for usage limits
  • Improve task descriptions with specific goals

Code Generation Issues

• Problem: Generated code has syntax errors
• Solution:
  • Verify all agents have role/goal defined
  • Ensure tasks have descriptions
  • Try simpler agent configurations first

---

📚 Learning Resources

CrewAI Documentation

• CrewAI Official Docs
• Agent Configuration Guide
• Task Definition Reference

Streamlit Resources

• Streamlit Documentation
• Session State Guide
• Custom Components

Groq API

• Groq Console
• API Reference
• Model Availability

---

🛣️ Roadmap

v1.1 (In Progress)

• [ ] Persistent workflow templates
• [ ] Agent performance metrics
• [ ] Task retry logic
• [ ] Workflow versioning

v1.2 (Planned)

• [ ] Database integration (save workflows)
• [ ] User authentication
• [ ] Team collaboration features
• [ ] Advanced scheduling

v2.0 (Future)

• [ ] Visual workflow diagram
• [ ] Agent skill marketplace
• [ ] Multi-LLM provider support (OpenAI, Anthropic, etc.)
• [ ] Production deployment guide

---

🤝 Contributing

We welcome contributions! Areas for improvement:

• Enhanced UI/UX components
• Additional workflow templates
• Performance optimizations
• Test coverage expansion
• Documentation improvements

Development Setup

git clone https://github.com/guiedup/Streamlit-Crew-AI.git
cd Streamlit-Crew-AI
pip install -r requirements-dev.txt
streamlit run app.py

---

📋 Requirements

| Package | Version | Purpose | |---------|---------|---------| | streamlit | >=1.28.0 | Web UI framework | | crewai | >=0.1.0 | Multi-agent orchestration | | langchain-groq | >=0.0.1 | Groq LLM integration | | python-dotenv | >=1.0.0 | Environment config |

See requirements.txt for complete list.

---

📄 License

MIT License - See LICENSE file for details

---

🙋 Support & Contact

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: Available in GitHub profile

---

⭐ Show Your Support

If this project helped you build amazing AI workflows, please:

• Star ⭐ this repository
• Share with the community
• Contribute improvements

---

🎓 About the Author

Guilherme Eduardo Pereira (@guiedup)

• Full Python Developer
• AI Automation & Computer Vision Specialist
• Mechatronics Engineer
• Building scalable automation solutions with LLMs and Python

Connect: GitHub | Website

---

Last Updated: 2026-04-13
Status: Production Ready ✅
Version: 1.0.0