Supported inputs
Canonical UDS conversion currently supports PDF, DOCX, XLSX, CSV, HTML, Markdown, text, PNG, JPEG, WebP and GIF. Other format pairs vary and unsupported paths fail visibly.
Migration Bug Light 9 · gradual adoption · source-preserving
You do not have to start over.
Your PDFs, Word files, text, Markdown and supported images can become the starting point for Universal Document adoption.
What works today?
What can survive
Extractable text, supported images/structure, reliable metadata, source hash and explicit conversion evidence.
What cannot be recreated
Missing revision history, certain authorship, identity, custody, approval and semantics are never invented.
Start with your responsibility
CIO / CTO
Modernize document boundaries without replacing every repository.
Open one-page brief →Enterprise architecture
Define a reversible source-to-UD layer with explicit exceptions.
Open one-page brief →Records management
Preserve original bytes, lineage, retention context and reconciliation.
Open one-page brief →General Counsel
Keep source evidence while making final state and later change inspectable.
Open one-page brief →Legal operations
Pilot one document family before changing the legal stack.
Open one-page brief →Compliance
Separate conversion evidence from compliance conclusions.
Open one-page brief →Information governance
Inventory, classify and reconcile rather than silently dropping files.
Open one-page brief →Healthcare HIM
Use de-identified samples and preserve source/derived distinctions.
Open one-page brief →Government records
Evaluate open, accessible outputs without claiming archival acceptance.
Open one-page brief →Tax operations
Preserve filing sources and amendment relationships without claiming authority acceptance.
Open one-page brief →Banking and financial operations
Preserve statement/report source evidence without claiming numerical correctness.
Open one-page brief →Libraries and archives
Keep original bytes and document what future readers need to interpret derivatives.
Open one-page brief →IT migration teams
Measure success, warnings, failures and manual review per batch.
Open one-page brief →Developers and integrators
Use explicit mappings, machine-readable outcomes and stable identifiers.
Open one-page brief →Migration authority cluster
You Do Not Have to Start Over
Adopt UD incrementally: create new UDs now and convert high-value legacy families when the benefit justifies it.
Read the answer →Conversion Should Never Invent History
A converter may extract source facts and infer structure, but must label inference and leave unavailable history unknown.
Read the answer →What Happens to Provenance When You Convert a PDF?
Preserve the exact source hash and available metadata; do not manufacture authorship, approval or custody evidence.
Read the answer →A Scan Can Become Structured Without Pretending It Was Born Structured
Keep the source image, OCR text, confidence and human corrections as distinct evidence layers.
Read the answer →What Happens When You Have 100,000 PDFs?
Inventory, sample, define a profile, validate exceptions, process staged batches and reconcile every source.
Read the answer →The Original File Should Remain Part of the Story
A derivative UD should reference the original format, exact hash, size, conversion time, tool version, warnings and result.
Read the answer →Migration Is Not the Same as Conversion
Conversion creates outputs; migration also governs inventory, validation, exceptions, reconciliation, rollout and long-term stewardship.
Read the answer →How Do You Move From PDF to a Verifiable Document Architecture?
Begin with a representative sample, preserve sources, quantify extraction limits, validate resulting UDs and expand only after acceptance.
Read the answer →Free individual conversion. Paid bulk migration.
Basic one-file conversion remains the low-friction adoption path. Bulk Document Conversion adds inventory analysis, mapping profiles, batch processing, exceptions, reconciliation, quality review and reporting. Existing pricing and fulfillment remain unchanged.