Two ways to get intake data into Best Case
Upload the intake you already have, or capture the client interview directly. Both routes lead to reviewed case data and a verified Best Case file.
Start with a packet or the interview
Upload the completed intake as a handwritten scan, typed PDF or page photos. If there is no packet, dictate live during the interview or paste the notes you already took. Manual schedule-by-schedule entry is available too.
Uploaded packets are read five times
For packet uploads, each read is independent and the five results are compared field by field. Agreement across all five reads is a strong signal that the value is right; disagreement is a signal that a person should look. Dictated and pasted notes skip OCR and move directly into structured review.
Your staff review the mapped information
For uploaded packets, fields the reads disagreed on appear beside the source scan. For dictated or pasted notes, staff review extracted and default values before generation. Your team corrects anything needed, and every choice is recorded against the case.
Confirmed values map onto Best Case schedules
Mapping is deterministic. Confirmed values land on their Best Case fields and schedules through fixed rules rather than a model deciding where things belong, so the same intake always produces the same file and every mapping can be inspected before export.
The exported .BCB is verified before you get it
Casewell writes a native Best Case file, then reads that file back and checks it against the values your team confirmed. A file that does not match is not handed over. What you import is what your firm approved.
What Casewell does not do
Being clear about the boundary matters more in legal software than in most categories, so it is stated plainly rather than left to be inferred.
Frequently asked questions
Can OCR read handwritten intake forms?
Yes, but a single OCR pass over handwriting is not reliable enough for a bankruptcy petition. Casewell handles this by reading the same page five times independently and comparing the reads field by field. Fields where every read agrees are treated as confirmed. Fields where the reads differ are shown to your staff next to the cropped handwriting from the original page, so a person decides rather than the software guessing.
Can OCR read cursive handwriting?
Cursive is the hardest case for any OCR system, and accuracy drops on it. That is precisely why Casewell reports agreement between reads instead of a single answer: cursive fields are the ones most likely to come back as a disagreement, which routes them to a human instead of quietly producing a wrong value.
What file formats can I upload?
Upload a PDF, PNG, JPEG or WebP file, including a handwritten scan, a typed or filled-in PDF, or page photos. Instead of uploading a packet, staff can also dictate live in a supported browser or paste interview notes.
How long does a case take to process?
Five reads across a full intake packet usually take a few minutes. The work runs in the background, so you can close the tab and come back to it; nothing depends on the browser staying open.
What happens if the reads disagree on a lot of fields?
You see them all, ranked, with the source handwriting beside each one. A packet with messy handwriting produces more confirmations and less time saved, but it never produces a silently wrong petition. If a particular form or handwriting style produces persistent disagreements, tell us and we will tune the read for that form.
Does Casewell remember corrections?
Yes. Corrections your firm makes are recorded against the case and used to improve how your own intake form is read. Your firm’s data is not used to train anyone else’s reads.