Dawrly a Crew of Ai agents

• Dawrly is a multi agent system designed to automate the job hunting process. (so yeah boy no need to spend a lot of time again on linkedin or indeedd , etc..)

• As you probably know, the current tech job market is pretty tough. After studying MAS (Multi-Agent Systems), I got the idea to build Dawrly, an AI job hunter that automates the entire process for us.

if you still got any question about the project

• you can check the deep wiki docs for this project and chat with Mr Devin here about it

  • 

• or speak with me I'd be more than happy to help

  • Linkedin
  • ziadwaleedmohamed2003@gmail.com

How Does it work ? 

• check this Demo and the Following diagrams

🎥 Demo Video (Click on the thumbnail)

Diagrams 

Agent Number 3 (Job scrutinizer) workflow in details:

• Cause it's a custom LangGraph workflow while the other 2 agents are built using CrewAI

Click to view the diagram

graph TD
    A["Start: Read Job URLs"] --> B["Scraping Node"];
    B --> C{"Scraping Successful?"};
    C -- Yes --> D["Filtering Node"];
    C -- No --> H["Drop / Skip URL"];
    D --> E{"Filtering Successful?"};
    E -- Yes --> F["LLM Analysis Node"];
    E -- No --> H;
    F --> G{"Matches Requirements?"};
    G -- No --> H;
    G -- Yes --> I["Schema Conversion"];
    I --> J["Collect Valid Jobs"];
    J --> K["End"];

Loading

How to use it ?

you have got two ways to try it out :

  1. the demo while it's live Click Me

  2. you can use it locally on your own device

For Local Installation 💻

The Docker Compose Way 🐳

• This method allows you to run Dawrly and its dependencies using Docker containers, providing an isolated and consistent environment.

---

1. Clone the Repository:

   First, you need to clone the project repository:

   git clone https://github.com/ZiadWaleed2003/Dawrly-Crew-of-AI-Agents.git
   cd Dawrly-Crew-of-AI-Agents

2. Ensure Docker and Docker Compose are Installed:

   Make sure you have Docker and Docker compose locally

3. Configure Environment Variables:

   Both the backend and frontend directories require a .env file for configuration.

   • Create .env files: Navigate into the backend directory and create a file named .env:

     cd backend
     touch .env

     Then, do the same for the frontend directory:

     cd ../frontend
     touch .env
     cd .. # Go back to the main project directory

   • Populate .env files: Open the newly created .env files in your text editor and add the required environment variables based on their respective .env.example files.

     Example: backend/.env.example

     AGENTOPS_API_KEY=your-api-key
     TAVILY_API_KEY=your-api-key
     FIRECRAWL_API_KEY=your-api-key
     CEREBRAS_API_KEY=your-api-key
     NVIDIA_API_KEY=your-api-key
     GEMINI_API_KEY=your-api-key
     GROQ_API_KEY=your-api-key
     LANGSMITH_API_KEY=your-api-key
     LANGSMITH_TRACING=true
     LANGSMITH_PROJECT=anyname-you-want
     EMAIL=email to use when sending the results
     EMAIL_PASSWORD= I guess this one is obvious

     Example: frontend/.env.example

     # Example variables for the frontend
     VITE_BACKEND_URL=http://localhost:8000

4. How to get the API keys ?

   • unfortunately while developing this app I used alot of services but the good news it's all free and you can get the API keys too :
   • Nvidia NIM
   • Agentops
   • Tavily
   • FireCrawl
   • Cerebras
   • Gemini API
   • Groq
   • LangSmith
   • Email --> you need to create a new google account and use it here for the app to be able to send you the results on your perosnal email
   • Email Password --> you will also need to get it from google

   add these to the .env file in the backend and you are good to go

5. Run with Docker Compose:

   Once Docker is installed and your .env files are configured, navigate to the main project directory (where your docker-compose.yml file is located) and run the following command:

   docker compose up --build

   This command will:

   • up: Start all services defined in docker-compose.yml.
   • --build: Rebuild images if there are any changes in Dockerfiles or context.

   After the services are up and running, you should be able to access the frontend of Dawrly in your web browser, at http://localhost:3000 (make sure to kill any process running on port 3000 or 8000).

What ? did u expect me to provide a pip install method ?

• I'm going to hold your hands when I say this (ma man you need to learn how to use Docker it's 2025 and you are still drowning in dependency conflicts)

6. I almost forgot ... for monitoring I used LangSmith and AgentOps

• Don't worry there is nothing more u need to do just add the env vars and you are good to go you will find the logs on LangSmith website and AgentOps