Applying our learning
Let’s use what we have learned to build a SQL chatbot using the Favorita project’s table metadata, monitor the streaming transaction project’s model, and deploy the chatbot that we have assembled and evaluated.
Technical requirements
The technical requirements needed to complete the hands-on examples in this chapter are as follows:
- The SQLbot will require OpenAI credentials.
- We will use the Databricks Secrets API to store our OpenAI credentials.
- You will need a personal access token (PAT) to deploy your web app to HF. See Further reading for detailed instructions.
Project: Favorita store sales
Let’s build a simple SQLbot using OpenAI’s GPT to ask questions about our Favorita Sales tables. Please note that while this section continues to use the Favorita Store Sales data, it is not a continuation of the earlier project work. In this example, you’ll create instructions on how the bot can...
