In the last month, we released several feature improvements, including the ability to match the most frequently repeated value for a repeated anchor and fuzzy matching for the Currency type.
"match": "mostFrequent" enum on the Field query object returns the most frequently matched value. This is useful for OCR text, like poor-quality scans or photographs. For example, if a scanned document contains repeated data for a field anchored on "1 Wages", but due to OCR errors the matched values are
21850.58, this option returns the most frequent, and therefore the mostly likely correct value,
The Name type now recognizes lists of names, for example
Argos Fullington, Jax Odenson, Ollie Longstreet.
We made several improvements to Sensible's ability to recognize addresses.
Use the new Relaxed With Cents parameter with the Currency type when poor-quality scans or photographed documents result in erroneous OCR output for the decimal separator or thousands separator.
If you set this parameter, Sensible overrides all other Currency type parameters and outputs USD currency.