Scoresheet Scanning
Turns re-typing a paper scoresheet into a photograph-and-review flow.
Photograph
A teacher photographs their paper scoresheet (or uploads an existing photo) from the exam’s score-entry screen.
Extract
A vision model reads the photo and extracts a table — student names/admission numbers and their scores per column — matched against the exam’s actual configured assessment components (so the columns line up with what the school actually grades on, not a generic guess).
Review
The extracted data pre-fills the same digital scoresheet a teacher would otherwise type into by hand. Nothing is uploaded yet — this is a draft for the teacher to check and correct.
Approve
The teacher reviews, fixes anything mis-read, and approves. Only then is the data written to the gradebook — through the exact same pipeline a manually-typed scoresheet goes through (student matching, column mapping, flags for anything that needs resolution).
If no vision provider is configured, this feature reports that clearly and refuses the request — it never falls back to fake or placeholder data. See Getting Your API Keys.