FinSight - A financial AI assistant tool for US-based companies

Requirements

1. Anaconda
2. Python 3.11+
3. Poetry
4. CrewAI
5. LangChain
6. Flask

Installation Steps

1. Add .env file in the root directory with all the API Keys

2. Clone the FinSight project

git clone https://github.com/Tamal-Mondal/FinSight.git
cd FinSight

3. Create a conda environment with Python 3.11

conda create -n crewai-finance python=3.11 -y
conda activate crewai-finance

4. Install all the depenedencies as per pyproject.toml

pip install poetry platformdirs flask flask-cors
poetry install --no-root

Usage

Summarizing 10K/10Q Reports

Mention a specific 10K/10Q report of any US based company, the model will return a brief summary of it.

End-point

https://localhost:5000/insights

Payload

{ 
    "company": "Oracle",
    "year": 2023,
    "form-type": "10Q",
    "quarter": "Q4"
}

Question-Answering on the basis of 10K/10Q reports

Based on the specific 10K/10Q reports, you can ask any question and should get a crisp answer.

End-point

https://localhost:5000/qna

Payload

{ 
    "company": "Microsoft",
    "year": 2023,
    "form-type": "10K",
    "quarter": "",
    "question": "How was the growth of the company in 2023?"
}

Stock Analysis and Prediction

Mention a company name for which you want stock related prediction. The model will analyze the company financials and advice you accordingly.

End-point

https://localhost:5000/analyse-stock

Payload

{ 
    "company": "Google"
}

Trends Visualization

Enter a company name, the model will return the past 10 years of data for some of the important metrics.

End-point

https://localhost:5000/visualization

Payload

{ 
    "company": "Google"
}