title	emoji	colorFrom	colorTo	sdk	app_port	license	short_description
Expressly	🛠️	blue	purple	docker	7860	mit	Expressly - Text Transformation App

Expressly - Text Transformation App

This document provides an overview of the backend server for the Expressly - Text Transformation App. The app is designed to transform text based on user preferences and descriptions, leveraging a CrewAI-based multi-agent AI application.

Key Features

Text Transformation Options

The backend supports the following transformation capabilities:

1. Formats

   • For details on supported formats, please refer to the Expressly Wiki.

2. Tones

   • For details on supported tones, please refer to the Expressly Wiki.

3. Target Audience

   • For details on supported target audiences, please refer to the Expressly Wiki.

Output

The backend processes user inputs and generates formatted text tailored to the specified preferences. Users can easily copy and utilize the output.

Backend Configuration

Prerequisites

1. Install uv and crewai if not already installed:

   pip install uv crewai

2. Navigate to your project directory and install dependencies:

   crewai install

   (Optional: Lock dependencies using the CLI command.)

Customization

1. Add environment variables to the .env file:

   MODEL=gemini/gemini-1.5-flash
   MODEL_API_KEY=<model_api_key> # Your API key here
   

   Generate a new API key for GEMINI model from the AI Studio website.

2. Modify configuration files as needed:

   • src/expressly_server/config/agents.yaml: Agents configuration.
   • src/expressly_server/config/tasks.yaml: Tasks configuration.
   • src/expressly_server/crew.py: Crew configuration with custom logic, tools, and arguments.
   • src/expressly_server/main.py: Entrypoint to kick off the crew with inputs for agents and tasks.
   • src/expressly_server/app.py: FastAPI configuration.
   • src/expressly_server/web_app.py: Gradio configuration.

Running the Backend

To start the backend server and execute tasks:

crewai run

Deployment

The app is deployed on Hugging Face Spaces. You can try it out here.

Docker

Check the docker documentation here. To run the docker image, use the following command:

docker run -p 7860:7860  -e MODEL="gemini/gemini-1.5-flash" -e MODEL_API_KEY="<model_api_key>" deepak93p/expressly-app

Additional Notes

• Ensure all environment variables are correctly set in the .env file check for .env.example file for reference.
• Regularly update your agents and tasks configuration to enhance functionality.
• Refer to the CrewAI documentation for advanced customizations.

Contributing

We love contributions! Whether you’ve spotted a bug or have an awesome idea, feel free to open an issue or submit a PR.

Support

If you like my work and want to support give a ⭐ to the repo.

Thank you so much ❤️.

License

See the LICENSE file for more details.