We’ve already wrote a bit on this new integration here.

Briefly, we’ve integrated into PRESENTA the GPT-3 Prompt you can use to generate text and other data for image generation purposes.

… prompt

The Prompt can be used freely but the only requirement is that the final output needs to be a JSON object. This is required since PRESENTA engine needs such data structure.

Let’s see how to write a prompt that return a JSON encoded response.

… example promp json 1

… example promp json 2

Let’s see how such generated information can be used within a PRESENTA template. Consider the following template that contains a couple of parameters:

… template

Now, suppose we want to generate both a text and author, let’s say, about a famous quote:

… prompt quote

You can refine the prompt in order to fine-tune the desired response:

… prompt not ok

If the prompt response is compatible, you can apply it to the template:

… prompt ok

Now, you can render the template to test the final result:

… render

If everything is ok, you can grab the URL, as usual, for your integration.

… mailchimp

Here you can see how I can generate a bunch of quote cards automatically:

… list of images