donbr / AIMS-CrewAI-Demo

A demonstration of CrewAI put together by AI Makerspace!

Using CrewAI for Financial Analysis - AI Makerspace!

---

Introduction:

In this repository we have built a multi-agent "crew" (powered by CrewAI) meant to generate AI-powered financial advisement reports.

NOTE: This example does not provide real financial advice, and should not be used to guide investment strategy

Running the Crew!

We'll explore how to run the crew in detail below!

Step 1: Cloning this Repository!

1. We'll want to start by cloning this repository, which can be done as follows:

git clone https://github.com/AI-Maker-Space/AIMS-CrewAI-Demo.git

2. Next, we'll cd into the newly cloned repository:

cd AIMS-CrewAI-Demo

Step 2: Initializing Our Environment Variables

We'll be using the following APIs to help us today:

1. OpenAI's Text Embedding 3 Small Embedding Model
2. Anthropic's Claude Sonnet 3.5
3. Sec API
4. SerpAPI

Once we've collected all our API keys, we can continue to add them to our .env file.

1. Next we'll create a new empty .env file to store our actual environment variables in.

cp .env.sample .env

2. We'll add the content and fill in our API keys in the newly created .env file.

Step 3: Install Dependencies through Poetry

We can install our dependencies straightforwardly using Poetry!

poetry install --no-root

Step 4: Run the Agent Crew!

All that's left to do now is run the crew, which we can do with:

python main.py

After which it will ask for a company and we're off!

Credits:

This repository is largely based on the example from the CrewAI creator @joaomdmoura and the original can be found here!