CrewAI Starter Agent

A simple yet powerful AI research crew built with CrewAI that leverages multiple specialized agents to discover and analyze groundbreaking technologies.

References and Acknowledgements

• CrewAI documentation
• CrewAI + Nebius
• CrewAI examples
• Nebius Token Factory documentation
• This example is contributed from Arindam200/awesome-ai-apps

Features

• 🔬 Specialized Research: Dedicated researcher agent focused on discovering groundbreaking technologies
• 🤖 Intelligent Analysis: Powered by Meta-Llama-3.1-70B-Instruct model for deep insights
• 📊 Structured Output: Well-defined tasks with clear expected outputs
• ⚡ Sequential Processing: Organized task execution for optimal results
• 💡 Customizable Crew: Easy to extend with additional agents and tasks

Prerequisites

• Nebius API key (get it from Nebius Token Factory)
• If running locally, python 3.10 or higher dev environment.

Tech Stack

• CrewAI agent framework
• Nebius AI for LLM inference

Task Structure

Tasks are defined with:

• Clear description
• Expected output format
• Assigned agent
• Sequential processing

Example Tasks

• "Identify the next big trend in AI"
• "Analyze emerging technologies in quantum computing"
• "Research breakthroughs in sustainable tech"
• "Investigate future of human-AI collaboration"
• "Explore cutting-edge developments in robotics"

Setup

The code can be run locally or on Google colab. Colab is recommended, as it doesn't need any setup.

Local env setup

1. Clone the repository:

git clone   https://github.com/nebius/token-factory-cookbook/
cd  agents/crewai-research-agent

2. Install dependencies:

if using uv package manager

uv sync

If using conda/pip

pip  install  -r requirements.txt

1. Create a .env file in the project root and add your Nebius API key:

NEBIUS_API_KEY=your_api_key_here

Video

🎥 Video tutorial

Code

agent.ipynb notebook can be run locally or Google Colab. - - run locally:
uv run --with jupyter jupyter lab agent.ipynb

agent.py python script can be run locally.

# using uv
uv run python agent.py

# or 
python agent.py

Dev Notes

How the uv project was created.

uv init .
uv add -r requirements.txt
uv add --dev ipykernel   # for jupyter kernel
uv sync

# create a kernel to use uv env within vscode
source  .venv/bin/activate
uv run python -m ipykernel install --user --name=$(basename $(pwd)) --display-name "$(basename $(pwd))"
jupyter kernelspec list  # verify kernel is successfully created