Search for an existing document that matches the criteria supplied
There are of course limits to what can be searched upon. For a document search to work, you must supply at a minimum:
The service will return a list of matching documents with confidence levels.
If you are the “owner” of the document - i.e. the same CustomerID and CustomerChildID (if relevant) - then the full details of the document will be returned, except for the contents of any attached scans. If you are not the owner of the document, then just the ID and confidence level is returned. You can still use this ID to retrieve any check results (see GET /document/{documentId}/checks)
Note: At this time, we cannot perform searches on document scans. But, you can supply extraData KVPs if they’re known. These will help double check search results with ambiguous results.
Customer ID issued by Frankie Financial. This will never change. Your API key, which is mapped to this identity, will change over time.
If, as a Frankie Customer, you are acting on behalf of your own customers, then you can populate this field with a Frankie-assigned ID.
Note: If using a CustomerChildID, you will also need a separate api_key for each child.
Any documents, checks, entities that are created when this field has been populated will now be tied to this CustomerID + CustomerChildID combination. Just as Customers cannot see data created by other Customers, so too a Customer’s Children will not be able to see each other’s data.
A Customer can see the documents/entities and checks of all their Children.
Open string that can be used to define the “channel” the request comes in from. It can potentially be used in routing and risk calculations upon request. Default values that can be used are:
Any alphanumeric string is supported though. Anything over 64 characters will be truncated.
The ISO-3166-alpha3 country code of the issuing national. Once set, this cannot be changed.
See https://en.wikipedia.org/wiki/List_of_ISO_3166_country_codes for more
Valid ID types
This document’s data was initially created from scanned and processed images. The value cannot be set manually and any attempt to do so will just be ignored.
Collection of one or more objects that describe scan(s) that need to be put through OCR or facial recognition. These should all be from the one ID document, such as front/back, or page 1, 2, 3, etc. You can upload multiple scans in a single call, or in multiple calls.
Note: if you do upload over multiple calls, make sure you include the documentId (see above), and indicate that this is happening with a “more_data” checkAction
When an ID document is created/uploaded, it is assigned a documentId. You’ll see this in a successful response or successfully accepted response. This can then be referenced in subsequent calls if you’re uploading more/updated data.
Current status of a document.
Set of key-value pairs that provide ID type-specific data. If updating an existing document, then existing values with the same name will be overwritten. New values will be added.
If this document is scanned through OCR or similar processes, then extracted data will be found here (Some may be used to populate other fields like idNumber and idExpiry as well)
The expiry date of the document (if known) in YYYY-MM-DD format.
The issued date of the document (if known) in YYYY-MM-DD format.
The ID number of the document (if known).
The sub-type of identity document. Very document specific.
If this document was originally populated from scanned data, then manually adjusted (e.g. if the scan’s results weren’t 100% correct or data was missing), then this will be set to true. The value cannot be set manually and any attempt to do so will just be ignored.
Regional variant of the ID (e.g. VIC drivers licence)
You should always use the local abbreviation for this. E.g.
The request was valid and able to be processed in some fashion. Returns a list of potentially matching documents or document references, along with a confidence level in the match.
Unique identifier for every request. Can be used for tracking down answers with technical support.
Uses the ULID format (a time-based, sortable UUID)
Note: this will be different for every request.
The list of (potentially) matching documents with confidence levels.
If you are the “owner” of the document - i.e. the same CustomerID and CustomerChildID (if relevant) - then the full details of the document will be returned, except for the contents of any attached scans. If you are not the owner of the document, then just the ID and confidence level is returned. You can still use this ID to retrieve any check results (see GET /document/{documentId}/checks)