In this video tutorial, we'll walk through how to create an earnings call assistant using OpenAI Embeddings and Completions API.
To do this, the steps we need to take include:
- Getting an earnings call transcript from Financial Modeling Prep
- Preparing the data, splitting it into subsections, and computing document embeddings
- Finding the most relevant document sections to the user's questions
- And using the GPT-3 Completions API to answer user questions.
The video also includes a Colab notebook that shows how to set up the necessary libraries, constants, and parameters and provides step-by-step instructions for each function. Lastly, we discus the limitations of the system and the need for fact-checking.
In the next section, we'll walk through how to take this earning call assistant out of Google Colab and create a web app using Streamlit.