Create document to text operation type

Created on 2 October 2024, about 1 year ago

Problem/Motivation

Currently we have services like Unstructured IO and ConvertAPI that can take documents as PDFs, Excel, Word etc and convert hem into text and sometimes markdown or HTML.

We should unify these services since they sometimes do rely on AI, so they become ambigious. This would also open up for an easy way of using local Python solutions for this.

Since different services takes different documents, they will not be entirely hotswappable, but still similar.

Steps to reproduce

Proposed resolution

Add an operation type interface called DocumentToText.
Add a method calles handlesDocumentType that will respond true or false for mimetype or extension name.
Add a method calles handlesOutput that will respond true or false for what outputs it handles.
Add into the model capability what output it handles.
Add an output type that can produce text, markdown and html.

Remaining tasks

User interface changes

API changes

Data model changes

✨ Feature request
Status

Active

Version

1.0

Component

AI Core module

Created by

πŸ‡©πŸ‡ͺGermany marcus_johansson

Live updates comments and jobs are added and updated live.
Sign in to follow issues

Comments & Activities

Production build 0.71.5 2024