Skip to content

FrankRank stands as a Python-based web application designed to revolutionize financial decision-making through the implementation of an objective chatbot powered by ChatGPT. This innovative approach aims to mitigate biases and emotional influences inherent in human decision-making processes related to finances.

Notifications You must be signed in to change notification settings

Micz26/FrankRank

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FrankRank - Financial Advisory Chatbot

Problem

Finances often depend on subjective opinions, influenced by preconceptions or the potential dishonesty of financial advisors, leading individuals to be exposed to excessive risks due to poor advice.

Solution

FrankRank - Python-based web application that addresses this issue by utilizing a chatbot, specifically ChatGPT.

FrankRank stands as a Python-based web application designed to revolutionize financial decision-making through the implementation of an objective chatbot powered by ChatGPT. This innovative approach aims to mitigate biases and emotional influences inherent in human decision-making processes related to finances. Application boasts a robustly implemented backend logic akin to ChatGPT, integrating sophisticated CRUD (Create, Read, Update, Delete) functionalities seamlessly orchestrated through Django's powerful ORM (Object-Relational Mapping) framework.

What did we implement?

  • Prompt Engineering: OpenAI API is utilized for prompt engineering.
  • Web Development: We created a website using the Django framework that stores customer data and acts as an interaction intermediary.
  • Stock Analysis Functions:
    • Determine the company's trend on the stock exchange.
    • Format the latest news on the stock exchange.
    • Display companies' biggest shareholders.
  • Forecast Function: We developed a function for a forecast of the company's price.
  • Login System:: Enhances user engagement by tailoring the chat experience based on signup responses. Customers can answer specific questions during signup, thus influencing the chat system's responses for a personalized and intuitive platform
  • Yahoo Finance: Utilizing Yahoo Finance, we can provide real-time data that can be leveraged by the chat system in accordance with user requirements.
  • Web Scraping: The application performs web scraping to gather information from financial magazines.

Video GIF

Backend Overwiev

  • User Authentication and Authorization:
    • Signup and login views enabling user registration and login with username, email, password, and OpenAI API key.
    • "Login with Google" option for streamlined access using the OpenAI API key.
  • User Settings Customization:
    • Settings page for user-specific preferences like investments, sectors, risk level, and other parameters to fine-tune the GPT-based advice system.
  • Chat Management:
    • Categorized chats covering personal finance, investments, insurance, and car insurance.
    • Dropdown menu for users to switch between chat categories.
    • Record of past chats, allowing users to view previous conversations.
    • Access to unique URLs for individual chat sessions for easy retrieval and continuation of discussions.
    • Ability for users to create new chats within the current category for initiating fresh discussions related to the chosen financial aspect.
  • Logout Functionality:
    • Secure logout to terminate user sessions and maintain account security.
  • Integration with OpenAI API:
    • Utilization of the OpenAI API to power the chatbot functionality for objective financial advice generation.
  • Backend Navigation and Redirection:
    • Redirects from signup to settings page for initial customization.
    • Seamless user experience by categorizing and displaying only chats relevant to the selected category.

Technologies Used

  • Django: A high-level Python web framework for building robust web applications.
  • OpenAI: The OpenAI API is employed for fine-tuning ChatGPT and integrating it into the chatbot.
  • Azure Storage Blob: The application utilizes Azure Storage Blob to store and manage data.
  • Semantic UI / HTML: Semantic UI and HTML are used for frontend development, ensuring a user-friendly interface.
  • Data Warehousing: We configured a Data Warehouse (DWH) on Azure SQL to store and manage customer data.

Getting Started

pip install -r requirements.txt
python manage.py makemigrations
python manage.py migrate
python manage.py runserver

About

FrankRank stands as a Python-based web application designed to revolutionize financial decision-making through the implementation of an objective chatbot powered by ChatGPT. This innovative approach aims to mitigate biases and emotional influences inherent in human decision-making processes related to finances.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published