I'm working on a project that regroups all best AI (AIaaS) from different providers (GCP, AWS, Azure, DeepL, etc.) in one API (https://github.com/edenai/edenai-apis).

I've got asked the question : why aren't you regrouping Open Source models (instead of proprietary APIs) into one repo?

Well because it doesn't make sens to deploy and maintain large pytorch (or other framework) AI models (especially for document parsing, image and video moderation or speech recognition) in every solution that wants AI capabilities. So using APIs makes way more sens.

Deployed OpenSource models are being included using different APIs like HuggingFace and other equivalents.

The current plan is to add some of these AI capabilities into n8n automation (https://github.com/n8n-io/n8n) and Odoo App (https://github.com/odoo/odoo) (receipt, resume and ID document parsing, content moderation and translation) .

Some other capabilities can include : summarization, keyword extraction, sentiment analysis for text data and also speech to text or image and video tagging.

Do you know other Open Source projects You want this could integrate to ?