An inventory management app where there's a capture and viewing mode.
This is inspired by Estee’s app MiniQR where there’s also a create & scan mode.
In capture mode, users can give a prompt for what is expected to be detected, for example, “Magic the Gathering trading cards”, and the possible fields that can be inferred, e.g. "name", "mana cost". Users can use the camera feature to scan for objects, or upload images to scan for objects.
Objects detected will have a scan frame around them. After picking up all the possible objects, they will be added into the db and be labelled with what they're identified by the computer vision model for metadata. Text in objects should also be recognized and classified like “card effects”, “card costs” and stored in metadata. If the user has given a list of possible fields that can be inferred, we should try to fit those too. These objects should also have their own individual images that's cropped by the scan frame to be stored in the db for display to user.
In a single capture mode session, the results should be stored in a folder, where the folder description is the prompt they gave. The title is an ai interpretation of a suitable folder name based on the prompt they gave.
In viewing mode, users can navigate to different folders. They can rename their folders and edit the folder description. They should be able to toggle between grid or list view. They should be able to see the quantity of each item detected which defaults to 1, but identical objects detected should show under the same object type and reflect under quantity. They should be able to see the source image for each object (the scan frame), and be able to edit any fields, such as the name, quantity, and other metadata like card description. They should be able to move, duplicate objects to different folders, and also delete objects.
CacawInventory
CacawInventory is an AI-powered inventory manager tailored for collectible enthusiasts. It helps users scan, catalog, and organize physical items such as trading cards, figures, and plushies with pixel-art charm and advanced tech like AI-based detection and image analysis. The product is local-first, with upcoming support for cloud sync, offline usage, and community sharing.