This video shows how to use a new AI feature in Google AppSheet that automatically extracts text from images like receipts, without needing to train a custom model. AppSheet has had an OCR feature for a while, but it required training on specific types of images. The new AI-powered option is much simpler and more flexible. Instead of training it, you just describe what text you want pulled from the image and it figures it out for you.
To demonstrate, I created a simple app for storing receipt images along with basic information about each one, like the company name, total amount, and date. When I add a new receipt by taking a photo and selecting a category, I can click a button called “Fill from Receipt” that uses the new AI extract feature. It takes a few seconds, but then automatically fills in all the key fields from the image — even if the text is in tricky spots or overlapped by graphics.
You can also automate this process by setting up a scheduled task to extract data in the background. The video goes over how to do that, including a breakdown of the data model and the automation logic.
There are two important notes: this feature is currently in preview, and it’s only available if you have at least one account with an enterprise plan. You can still use the app on other accounts once it’s built with an enterprise one.
If you found this video helpful and you’d like to be notified when I post new videos about AppSheet, please subscribe to the channel.
Leave a Reply