Skip to Content
Assistant Building
Scan image to auto-fill form
Mavenoid's vision assist allows users to enter information like serial numbers by just snapping a photo of their product label rather than having to type everything in by hand. Our AI can scan this image, recognize the text, and categorize things like brand names, model numbers, and serial numbers to fill in form fields automatically.
To set this up in your flows, follow our step-by-step guide.
1
Add a scan image form field
In the node's edit pane, click the orange plus icon in the Form section and select "Scan image".
This will add a button to the node's form that users can click to upload a photo to be scanned.
Give the button a descriptive label, such as "Scan product label to auto-fill".
2
Add the fields to auto-fill
For each piece of information to be extracted from the image, such as the model number or serial number, add a text field.
Specify a form data key to use, a label to display, and any other options as appropriate.
3
Configure auto-fill
For each field that should be auto-filled from the image, make sure the "Enable auto-fill with scan image" box is checked.
This will add the "Category" dropdown. Pick the option that best matches the type of information this field should contain; this will guide the AI to pick the right piece of information from the product label for each field.
4
Preview to test
Preview the node to test the auto-fill. Click the scan image and upload an image containing a product label and confirm the fields are filled in correctly.
Product labels can vary, so make sure the feature works correctly with your products' labels before pushing this live to users. If it doesn't seem to work correctly with your labels, contact [email protected] for assistance.