The everyday jobs are covered without exception. Clean scanned PDF straightens and lifts faded faxes; OCR makes them searchable in eleven languages; Redact re-renders pages so a patient’s identifiers are removed rather than covered; Fill forms and Flatten handle consent and referral paperwork; Merge and Protect prepare a record for a controlled handoff. The app installs from the browser and keeps working offline once loaded, which suits a consulting room with an unreliable connection as well as a policy of processing patient data on-premises.
The jobs, and the tool for each
Clean up a faxed or photographed referral
Skew, weak contrast and dust are fixed on the device so the record is readable and ready for OCR. Keep the untouched original for the chart; cleanup changes pixels.
Clean Scanned PDF →Redact patient identifiers before releasing a record
Marked pages are re-rendered with the boxes burned in, so a name or a date of birth is removed from the file rather than hidden under a rectangle. Search lists catch every occurrence, and a redaction log documents what was done.
Redact PDF →Make intake forms and old charts searchable
The OCR engine runs in the browser in eleven languages; scanned pages are never sent to a cloud recognition service. Spot-check names and dosages — recognition is not transcription.
OCR PDF →Fill consent and referral forms on screen
Text fields, checkboxes, dropdowns and radio groups are completed in the browser and the form is never uploaded. Flatten afterwards so a completed consent cannot be altered by accident.
Fill PDF Forms →Flatten a completed form before it goes in the chart
Answers become permanent page content and stray annotations are removed, so what was signed is what stays on file. Flattening is not encryption; add Protect for a copy leaving the practice.
Flatten PDF →Merge a record and lock it for a controlled handoff
Reports, results and correspondence combine on a visual board in the order you choose, with a real outline. Password-protect the result with Protect before it travels.
Merge PDF →What leaves your device
Documents never reach Bindery. There is no upload endpoint for PDFs, so there is no protected health information on Bindery’s servers for Bindery to protect, and consequently no business associate agreement for Bindery to sign — the record stays on the practice’s device throughout. That is an architectural fact, not a compliance certification: Bindery does not claim HIPAA compliance or any certification, and the practice’s own obligations for the devices, browsers and storage it uses are unchanged. An optional account stores an email address, sign-in identity, plan status, a minimal daily job count and any saved recipes or presets — settings, not documents. The two features that do use Bindery’s storage, Secure handoff links and signature requests, receive only packages sealed on the device with AES-256, plus token and SHA-256 hashes; Bindery never holds the key. Teams records a count of jobs per member with no file names, tool names or sizes.
The full list of services, retention periods and subprocessors is on the security page; the verify page shows how to confirm the no-upload claim in your own browser.
Questions people ask
Does Bindery sign a BAA?
There is nothing to cover: patient documents are not transmitted to Bindery, so Bindery is not a business associate that stores, receives or processes PHI on the practice’s behalf. The account layer holds an email address and plan data only. Confirm this against your own compliance requirements rather than taking any tool’s word for it.
Can OCR run with the network disconnected?
Yes. The recognition engine and its language packs ship with the app and load from Bindery itself; once the app has loaded, OCR, redaction and every other tool keep working offline. That is also a simple way to verify for yourself that pages are not being sent anywhere.
What about redacting a scanned record that has no text layer?
Redaction works on the pixels: marked regions are burned into a re-rendered page image, so a scanned name disappears whether or not a text layer existed. If the scan had been OCR’d earlier, that hidden layer is removed on the redacted pages too, which is the failure other tools miss.