Parsing an array from glide to use as a chatGPT prompt

I have a knowledge base created in Glide which presents in as an array.

We have an extensive community using telegram, and I’m trying to collect Telegram questions and get ChatGPT to answer them using primarily our knowledge base. A great answer would be a short summary and then a link to 1-3 of the relevant articles.

So the process is:

  1. Collect telegram message
  2. Collect glide table rows
  3. Send a prompt to openAI saying “Please answer QUESTION by referring to DATA”
  4. Send the response back to Telegram

I have successfully connected my front and back of my scenario - here is my flow

Any advice how best to do this - both generally and specifically

Specifically my problem with this method is taking the glide output and turning it into something useful to pass to ChatGPT

This is my glide output

I’m also concerned this will be incredibly operation intensive in make and credit intensive in GPT so alternative methods appreciated!


Welcome to the Make community!

Your Glide output is already in an array, so you don’t need an aggregator to aggregate to an array.

In module 19, you can map something like this:

{{ join(map(9.rows.8VNvX); newline) }}

Otherwise, you can use an Iterator on the array 9.rows, and then use a Text Aggregator to combine them into a single string variable.

For more information, see

Here are some useful links and guides you can use to learn more on how to use the Make platform, apps, and app modules. I found these useful when I was learning Make, and hope they might benefit you too —


Help Center Basics

Articles & Videos

samliewrequest private consultation

Join the Make unofficial Discord server!

1 Like

Thanks! I will try that later. Is a long string the best way to feed ChatGPT? I’d ideally have it read the body and then reference the subject when providing a link (which is formed by one of the strings)- will this information be passed in a text string?

Well, you can create custom assistants and have the assistant refer to the uploaded files.

However now there is a “transition” period where OpenAI just released a new files handling method via a vector store and Make has yet to update their modules.

For more information, see My OpenAI assistant via make don't use the existing uploaded files - #2 by samliew


that is awesome! Don’t suppose you fancy sharing that flow in higher resolution?