Storing JSON data inside weaviate vector db

Hello everyone,

I was able to write a python script to add json data from a API feed. This stores data on the latest Hockey scores and matches.

After completing this, how would I get my openai APP to connect to weaviate, read the json data, and then be able to answer questions from this data?

For example, when is the next Penguins game? Did the Penguins lose last night?

Is this even possible? Would openai be able to take user queries (questions) and be able to read the json data and make sense of it all?

I thought it would but the questions are not being answered :frowning: Just responds normal gpt responses. Since the JSON data does have hockey scheduling and game info, I figured it would be able to answer these questions, but it seems like there is a bigger part to this.

Anyone can suggest what I am missing? Or which route to go? Do I need to store this JSON data inside a SQL db and then maybe use Text to SQL features? That means…I won’t have any need for weaviate?



Weaviate will connect to OpenAi for vectorizing the data and generating the answer, not the other way around.

Here we have a recipe that will show you exactly how to accomplish that:

Let me know if that helps!

Thank you. I will go ahead and take a look at this.