Three steps, from the job site.
It works from a phone. No new hardware and no retyping.
Photograph the form
Snap the paper form on site. Ovverlay reads it against that form's own rules.
Review the flags
You see only what needs attention: the field, the rule it broke and a plain-language note.
Submit for review
Group a job's forms and send a frozen snapshot to supervisors, who approve or send it back with notes.
Built so people can trust the result.
A flag should be something you can check, not a black box.
The model reads, code decides
The AI transcribes what is written. Whether two forms disagree is decided by plain, testable code.
Cross-form checks
Compare every form in a job against the source record, so a mismatch is caught once, not discovered later.
Private until submitted
A worker's photos stay private until they submit. Submissions are frozen snapshots, and review history is never overwritten.
Honest about uncertainty
Blank or unclear entries are skipped, and lower-confidence checks are labelled as advisory.
Starting in rail. Useful anywhere paper still rules.
Ovverlay was built around roadway-worker safety paperwork. The same approach fits any trade that runs on handwritten forms, such as utilities, construction and inspections.
Want to see it on your forms?
Tell us what paperwork your crews fill out and we will show you what Ovverlay would flag.