Trust Center
Why Trust CaseReady
CaseReady is built for people carrying court, CPS, family, housing, education, and benefits records. Trust here means you can see what is known, what is not, and who owns the files.
Why CaseReady is different
- Ask is limited to the current case. When a model is configured, it is instructed to cite your documents. Without a model key, answers are built from stored excerpts only.
- Every fact is labeled document verified, user reported, system generated, or unknown
- Private storage paths and row-level security scoped to your account
- You own the records; export and deletion are in Settings
- No claim of legal representation or guaranteed outcomes
The Trust Engine
CaseReady does not treat all text as equal. Briefings, timelines, and summaries should show how a statement was produced so you can check it.
Document Verified
Information supported by an uploaded PDF, with a file name and page when available.
Example source: Court Order.pdf · Page 4
User Reported
Information you entered. CaseReady has not independently verified it from a document.
Example source: Case journal entry
System Generated
Derived through processing, such as an upload event or an extraction pass. Confirm it against the source file.
Example source: Document added: School Records.pdf
Unknown
CaseReady cannot verify the claim from the records currently stored.
Example source: Not found in available records
How to verify
- Open the cited document from Documents.
- Go to the listed page when a page number is present.
- Compare the excerpt to the PDF. If it does not match, treat it as unverified.
- User-reported notes are yours to confirm with an attorney or original papers.
Private storage and security controls
1
Your browser (HTTPS)
2
CaseReady app (signed-in session)
3
Row-level security
4
Private storage + Postgres
Files are stored as user / case / document. Downloads go through an authenticated app route. The browser never receives a service-role key.
Transparent AI
Ask loads the case type, state, and county, then retrieves records from that case only. Uploading a PDF does not send the file to OpenAI. If a claim is not in the stored records, the honest answer is that it cannot be verified. Read the AI policy for limits.
