A user uploads an asset to be imported, and the OCR tool should hook into the database to view those assets (This might be a better way than viewing S3 assets). When finished OCR'ing a recipe, it should update the database to reflect the import is done, and connect all the assets together:
- Original asset
- Final Recipe
The user can view the original asset and double check the recipe.