Marketing Media Planner Project

This repository contains a project to provide marketing strategy using the key input of website name.

Prerequisites

• Python 3.11 or higher (required by langmem)
• Git

Installation

1. Set up a Python environment (Optional)

It's recommended to use a virtual environment to avoid package conflicts.

# Create a virtual environment
python -m venv venv

# Activate the virtual environment
# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate

# Verify Python version (should be 3.11 or higher)
python --version

2. Configure the environment variables

Create a .env file in the root directory of the project based on the provided example:

cp .env.sample .env

Edit the .env file and fill in the required variables according to your needs.

API Keys Configuration (Optional)

You can configure either OpenAI or Anthropic API keys in your .env file:

# Choose one of the following API configurations
# OpenAI configuration
OPENAI_API_KEY=your_openai_api_key_here

# Anthropic configuration
ANTHROPIC_API_KEY=your_anthropic_api_key_here

3. Install the package

Install the package in development mode:

python -m pip install -e .

This will install the package and its dependencies.

Running the application

To start the development server:

langgraph dev

This will start the langgraph development server, allowing you to interact with the application.

Accessing the Agent Interface

Once your LangGraph server is running, you can access the agent interface by visiting:

https://agentchat.vercel.app/

When you first visit the site, you'll need to provide:

1. Deployment URL - Your LangGraph server URL (e.g., http://127.0.0.1:2024)
2. Assistant / Graph ID - The name of your agent (typically agent)

No thread ID is required as the interface will handle conversation threads automatically.

Troubleshooting

• If you encounter any issues with Python version, ensure you're using Python 3.11 or higher.
• Make sure all the environment variables in the .env file are correctly set.
• For dependency issues, try removing the virtual environment and creating a fresh one.