Step-by-step guide on how to build AI agents using CrewAI

Full tutorial - https://www.youtube.com/watch?v=aOLXkdz3Wow

Creating AI agents using CrewAI is a straightforward process that allows individuals to automate various tasks without any coding knowledge. Here's a step-by-step tutorial to guide you through the process of building your own AI agents using CrewAI, focusing on automating content creation tasks as an example.

In this tutorial, we will build an AI crew with 2 ai agents. A researcher agent is responsible for conducting a comprehensive analysis of AI advancements and a writer agent is tasked with creating engaging blog posts based on the researcher's insights.

Our researcher here uses the DuckDuckGo search engine as a tool to gather information from the internet.

Step 1: Setup Your Environment

- Start with any IDE tool

Step 2: Install Necessary Libraries

- Install CrewAI and Other Libraries: In the first cell of your Colab notebook, type and run the following commands to install CrewAI and any other necessary libraries like DuckDuckGo search, which can be used by your agents for performing tasks.

pip install crewai
pip install duckduckgo-search         

Step 3: Setup Your CrewAI

- Import Required Modules: Import the necessary modules from CrewAI and other libraries you've installed.

from langchain.llms import OpenAI
from langchain_community.tools import DuckDuckGoSearchRun
search_tool = DuckDuckGoSearchRun()

import os
from crewai import Agent, Task, Crew, Process        

- Set Up API Keys: Configure your API keys for accessing models and tools. For example, if you're using GPT-3.5, you'll need an OpenAI API key. Go to the OpenAI website, navigate to the API section, and follow the instructions to generate a new secret key.

Step 4: Define Your Agents and Their Tasks

- Create Your Agents: Define the roles and goals of your agents. For content creation, you might have a 'Researcher' agent for gathering information and a 'Writer' agent for drafting blog posts.

- Create Tasks for Each Agent: Define specific tasks for your agents, detailing what each agent needs to accomplish. For instance, the researcher's task could be to find the latest AI advancements, while the writer's task could be to use those findings to write a blog post.

 # Define your agents with roles and goals
researcher = Agent(
  role='Senior Research Analyst',
  goal='Uncover cutting-edge developments in AI and data science',
  backstory="""You work at a leading tech think tank.
  Your expertise lies in identifying emerging trends.
  You have a knack for dissecting complex data and presenting actionable insights.""",
  verbose=True,
  allow_delegation=False,
  tools=[search_tool]

)
writer = Agent(
  role='Tech Content Strategist',
  goal='Craft compelling content on tech advancements',
  backstory="""You are a renowned Content Strategist, known for your insightful and engaging articles.
  You transform complex concepts into compelling narratives.""",
  verbose=True,
  allow_delegation=True
)

# Create tasks for your agents
task1 = Task(
  description="""Conduct a comprehensive analysis of the latest advancements in AI in 2024.
  Identify key trends, breakthrough technologies, and potential industry impacts.""",
  expected_output="Full analysis report in bullet points",
  agent=researcher
)

task2 = Task(
  description="""Using the insights provided, develop an engaging blog
  post that highlights the most significant AI advancements.
  Your post should be informative yet accessible, catering to a tech-savvy audience.
  Make it sound cool, avoid complex words so it doesn't sound like AI.""",
  expected_output="Full blog post of at least 4 paragraphs",
  agent=writer
)        

- Organize the Workflow: Decide on the sequence in which tasks should be executed (sequential or hierarchical) and create a Crew to manage this process.

# Instantiate your crew with a sequential process
crew = Crew(
  agents=[researcher, writer],
  tasks=[task1, task2],
  verbose=2, # You can set it to 1 or 2 to different logging levels
)        

Step 5: Execute and Monitor Your Agents

- Kick-Off the Process: Start the process and monitor the execution. You can print the results or outcomes of each agent's task to ensure everything is running as expected.

# Get your crew to work!
result = crew.kickoff()

print("###########")
print(result)        

Step 7: Review and Adjust

- Analyze the Output: Review the results produced by your agents. This might include the research findings and the drafted blog post.

- Iterate and Improve: Based on the output, you may want to refine the tasks, adjust the agents' tools or models, or reorganize the workflow for better results.

Conclusion

Congratulations! You've just learned how to create AI agents using CrewAI to automate the content creation process. This framework is flexible and can be adapted to automate a wide range of tasks beyond content creation, such as data analysis, customer support, and more. Experiment with different tasks, agents, and tools to explore the full capabilities of CrewAI in automating your workflows.