Base64ai
Base64ai
Aktiviteler
Data Extraction
This activity extracts structured information from documents, forms, invoices, tables, or free-text sources using predefined extraction models. The output is returned as a collection and can be used in downstream automation steps.
Usage Scenarios
Reading invoice fields such as vendor name, date, amount, tax
Extracting customer, contract or application form data
Converting unstructured text into structured variables
Validating and storing document information in databases
Automating back-office data entry from scanned files
Parameters
Document Path: Full file path of the document to process
Document Path Choose: Manual file selection from device
Document Text: Raw text to extract data from (optional)
Template Name: Name of the extraction template or model
Fields To Extract: Target field names separated by “;”
Confidence Threshold: Minimum accepted recognition score
Language: Document language (e.g., “en”, “tr”)
Delay Time: Waiting time before execution (milliseconds)
Notes
Document Path or Document Text must be provided
Fields To Extract must match the template field names
Low-quality scans or rotated images may reduce accuracy
Output is returned as a collection and should be assigned to a variable
Raw OCR
This activity extracts all textual content from an image or PDF using the Base64.ai service and returns the result as a collection. It is suitable for unstructured documents such as invoices, scanned forms, or text-containing images.
Usage Scenarios
Extracting full text from scanned invoices, receipts, or PDFs
Reading unstructured document content for later processing
Retrieving text from images used in reporting or validation
Converting visual data into searchable text collections
Parameters
Email: Registered Base64.ai account email
Secret Key: API access key obtained from Base64.ai
Image Or Pdf: File picker to select an image or PDF
Image Or Pdf Path Text: Manually entered image/PDF file path
Image Or Pdf Path URL: URL of an online-hosted document
Delay Time: Waiting duration before execution (milliseconds)
Notes
Use only one of: Image Or Pdf, Image Or Pdf Path Text, or Image Or Pdf Path URL
Output collection contains raw text lines or text blocks extracted from the document
Suitable for documents without a structured layout or defined fields
Vaccination Card Validation
This activity compares a vaccination card image with an ID image to verify their consistency and extracts vaccination details as a collection using the Base64.ai service.
Usage Scenarios
Validating whether a vaccination card belongs to the correct person
Extracting vaccine type, dose count, and date information for processing
Automating document verification workflows in healthcare or HR processes
Checking document accuracy before record submission or registration
Parameters
Email: Base64.ai account email
Secret Key: API access key from Base64.ai
Id Image: Manually selected ID image file
Id Image Path Text: Full file path of the ID image
Vaccine Card Image: Manually selected vaccination card image file
Vaccine Card Image Path Text: Full file path of the vaccination card image
Delay Time: Waiting duration before execution (milliseconds)
Notes
Both ID and vaccination card images must be provided
Images must be readable and belong to the same individual
Output collection includes vaccination details such as type, date, and dose count
Processing occurs on the cloud, so screen visibility is not required
