Everything Proof does, in the order you’ll use it.

Point Proof at a website and it finds every linked PDF: the crawl respects robots.txt, paces its requests to each site, and picks up where it left off after an interruption. Findings are reviewed, then imported. Most tools wait for you to upload files — Proof goes and finds them.

Staged multi-file dropzone: drop files, set Collection, tags and language, upload. Multi-file drops auto-create a Collection; exact duplicates are deduplicated by content hash.

Paste one URL or a list. Files are fetched safely in the background (HTTPS only, size limits, no requests to private networks), with per-item status that survives reloads, retry and cancel. Language auto-detected.

Upload a .zip of PDFs — up to 500 MB / 1,000 files — extracted and run through the same ingest path.

Priority lanes keep interactive work responsive while a 3,000-document import churns; a completion email tells you when the corpus is fully analyzed.

Every import — crawl, upload, URL, zip — lands in one hub inside the Library with per-batch status. Documents you already hold are re-checked against their source and refreshed when the published copy changes.

Every document checked against a WCAG 2.1 AA rule profile built on veraPDF, the open-source PDF/UA validator, plus Ockem’s own detectors for the checks it has no clause for. Runs at import and after every fix.

Image crops of the offending figure, extracted OCR text, per-page reading-order flags, per-run color-pair analysis. You always see why a rule fired — every check Proof asks a person to make points at the exact spot on the page, including language of parts and non-text contrast.

Compliant (green), Needs review (amber), Failed (red — confirmed violations only). A document awaiting human review is never called failed.

The contrast check sets aside text hidden behind full-page artwork — a kind of false alarm other checkers hand you raw.

Long operations stream page-by-page progress — “Tagging page 12 of 84” — as one continuous task.

Exposure KPIs, remediation-path lanes, a live effort calculator (hours → FTE-weeks), per-criterion WCAG scorecard, analytics CSV. A Certification & Evidence Register lists every certificate and who attested it, and a Remediation Ledger records what was done to each document — both exportable as CSV.

One picture is one question: a vision model drafts a short alt and an extended description in one place, with the figure crop beside the draft; a human approves or edits before anything is written. A picture that is really text is answered “It’s text” and transcribed. Describe an image once, and wherever it appears in your other documents Proof offers your approved description as a draft to review.

LLM-drafted table summaries, and a document title drafted from page 1 when the title is missing or just echoes the filename — human-reviewed, always.

Models run self-hosted on hardware we operate. No customer document content goes to a third-party model, and documents in your workspace train our models only if you turn that on — a decisive answer for government and healthcare security reviews.

Every change to the automatic tagger is measured against a benchmark and ships only if it makes nothing worse for compliance or safety. One version was built and deliberately not shipped because the test caught a regression.

Every document moves through an explicit status flow — claimed, in review, escalated, approved — with the approve action gated so a document sitting in amber or red can never be signed off. The gate is the product.

Contrast decisions, reading-order review, figure scan-confirm-dismiss, AI-draft accept/edit/reject, and human attestation each run as a dedicated queue — sorted by completion leverage (“N documents are one check from Compliant”), keyboard-driven, with a live count of what’s left.

Pull a document to yourself to work it, escalate an unrecoverable source for re-authoring instead of looping forever, and approve only once every check — machine and human — is resolved.

Per-document comments and a full status history record who did what and when — the review conversation and the audit trail live on the document, not in a folder of emails.

Configurable review gates enforced at the approval transition — org-wide or overridden per folder — plus teams with per-user library access, managed from the Tenant Admin console.

Assign documents with a due date and a hand-off note, and mention a teammate. Finished work is handed in for approval; an approver — who needn’t be an administrator — approves it or sends it back with a reason.

Routing rules send newly imported documents to a person by folder, tag or language, and assigned work leads each member’s feed.

Follow a document or a whole folder and get one daily digest of what changed; escalations reach a person. When someone leaves, their open work is handed on instead of orphaned.

Every query is org-scoped in the application and enforced by PostgreSQL Row-Level Security underneath. If the organization can’t be confirmed, access is refused, and automated tests that must pass before every release cover it.

One global identity, multiple organization memberships, on team plans Google SSO or sign-in through your own identity provider (OIDC) with mapping rules that place each member in the right team and project, MFA asserted by your provider honored, optional TOTP two-factor, a workspace picker for people who belong to several orgs.

Self-service admin for org owners: people and roles, teams and access grants, workflow review gates, security policy, audit activity, webhooks, notification defaults for new teammates, and an email delivery log.

Org-prefixed object storage, presigned expiring download links, structured logging, nightly backups. Model training on your documents is off unless a workspace administrator turns it on. Trial documents and the sample a free assessment scores are deleted after 30 days; a full import prepared for an assessment is deleted 90 days after its last activity.

Bring your agencies under one contract: one project per agency, with its own people, its own library, imports and crawls, and project-level access at read, work or admin. A Home page shows what is running now and finds any document across every project you can open; each project has its own address, and each browser tab is scoped to one tenant by URL.

Review and decide from a phone or tablet: the Library, the review queues, and each document’s Steps and Page tabs. Proof’s main pages reflow at 320 px wide with no sideways scrolling (WCAG 1.4.10); drawing tools ask for a larger screen.

Send certification, escalation, assignment, approval, import and remediation events to your own systems. Each endpoint shows its health; after ten failed deliveries in a row it is switched off and your admins are emailed.

A document-pack purchase provisions the organization, admin identity, and prepaid document balance — no sales call required. Admins are emailed when the balance runs low and again at zero.

Every file is checked for free first. A file that needs far more work than its page count suggests shows a fixed price in documents before anything is spent — accept one, accept several at once, or skip.