Brandkit - auto-tags & auto-descriptions via OpenAI API?

We’ve been experimenting with OpenAI’s API to see what’s possible and what makes sense for Brandkit customers. … and it’s amazing 👀 😮

It blows the current AWS Image Rekognition we’re using into the weeds.

We can now set custom prompts in account settings, with things like brand tone of voice, and get “on-brand” results.

Example 1:

Prompt Settings 1

“…Use formal and scientific language…”

Description 1:

“Photo of a Common Kingfisher (Alcedo atthis) perched on a branch; vivid orange and turquoise plumage against a soft-focus green background.”

Example 2:

Exact same image

Prompt Settings 2:

“…Use humorous and irreverent language…”

Description 2:

“Photo of a smug-looking kingfisher rocking a vibrant blue-orange feather coat, perching like it owns the twig runway. Background’s a blurry, envy-green.”

When, how?

We plan to release this to Brandkit® (OpenAI tagging, OpenAI descriptions) plus some other toys in the new year (Q1 2024). It’ll just turn up in your Admin console one day - but we’ll let you know more in due course.

Happy branding (and Merry Christmas) :)