MLQ Academy: Building a GPT-3 Enabled App with Streamlit

In this video tutorial, we'll be creating a simple web app using Streamlit that allows users to upload a PDF, input a YouTube video, or retrieve an earnings call transcript, and compute the embeddings of the document or video.

Users can then ask questions and get summaries of the documents using OpenAI's Embeddigns & Completions API.

The video provides step-by-step instructions to create the:

• Create an app.py file for Streamlit
• Create a menu that allows users to choose between a PDF assistant, YouTube video assistant, or earnings call assistant
• Create a file upload and YouTube video component and allow users to retrieve earnings call transcripts
• Use our OpenAI helper functions to compute embeddings for each use case
• Allow users to ask questions about the document or videos and display the responses

We'll be using the OpenAI API, PyPDF2, PTtube, whisper, and Financial Modeling Prep API to create the app.

The goal of this video and code is to provide a starting foundation for building apps with the OpenAI API and highlights the power and flexibility of Streamlit in creating web apps with minimal code. By following the step-by-step instructions, you can quickly create a functional web app that can be easily expanded upon.