LangGraph + CrewAI - Crash Course for Beginners [Source Code Included]

---

---

AI Nuggets

Based on the provided transcript, here are the detailed instructions, CLI commands, website URLs, and tips extracted in an easy-to-follow outline form:

Setting Up Email Automation Tool with Crew AI and LRA

Prerequisites:

• Gmail account
• Python environment

Tools Used:

• Crew AI
• LRA (Lang graph)

Steps:

1. Create Gmail API Credentials:
   • Visit Google Cloud Platform and create a new project.
   • Navigate to APIs & Services > Credentials.
   • Click Create credentials and select OAuth client ID.
   • Choose Desktop app as the application type.
   • Name your client (e.g., Crew AI Lang Graph).
   • Download the credentials.json file and rename it to credentials.json in your project directory.
   • Add credentials.json to your .gitignore file to avoid pushing it to version control.
2. Enable Gmail API:
   • Search for Gmail API in the Google Cloud Platform.
   • Enable the Gmail API for your project.
3. Set Up Environment Variables:
   • Create an .env file with the following variables:
     • EMAIL - Your Gmail address.
     • TALI_API_KEY - API key from Tali for internet searches.
     • OPENAI_API_KEY - API key from OpenAI.
4. Install Dependencies:
   • Use conda to create a new Python environment:

     conda create --name crew_ai_lang_graph python=3.11  

   • Activate the new environment:

     conda activate crew_ai_lang_graph  

   • Install required packages from requirements.txt:

     pip install -r requirements.txt  

5. Run the Tool:
   • Execute the main script to start the email automation tool:

     python main.py  

   • The first time you run the script, you’ll need to authorize the application to access your Gmail account.

Tips:

• The tool will connect to your Gmail account, pull down the latest emails, and generate draft responses.
• It will filter out junk and promotions, focusing only on real people’s emails.
• The tool uses Crew AI to analyze the text and generate responses.
• LRA is used to intelligently trigger the Crew AI based on certain conditions.
• The tool will create a graph with nodes representing actions and edges representing connections between actions.
• State is used to remember what has been done, such as which emails have been checked or which drafts have been written.
• The tool will loop, waiting for new emails and generating responses as they arrive.

Additional Resources:

• Free school community for AI developers: AI Developer Accelerator (no URL provided).
• Crew AI tools demo and source code (no URL provided, but mentioned to be available below the video).

Remember to replace placeholder text (e.g., Your Gmail address, API key from Tali, API key from OpenAI) with your actual information.