SOURCE DOCUMENT PREFLIGHT

A readable PDF.
A missing piece.
Find the gap.

What looks clear on a page can get lost during AI ingestion. Check text layers, tables, images, and layout before you send your document downstream. Supports PDF, DOCX, PPTX, and XLSX.

Check your document Explore a sample report

No account needed · No LLM key · Source stays unchanged

FROM SOURCE TO SIGNAL

EXAMPLE REPORTmixed.pdf

78/ 100
Worth reviewing

Likely scanned pagePage 2 · no extractable text

Image coverage100%

See the evidence

A generated example. A heuristic score, not an accuracy rating.

LOOK BENEATH THE PAGE

Text layers

Scans & images

Reading order

Tables

Hidden context

WHAT COULD GO MISSING?

The file opens.
The questions start there.

A PDF can look complete and still be difficult to ingest. Preflight surfaces measurable signs that deserve a closer look.

01 / TEXT

A page you can see,
but cannot extract.

Find image-dominated pages with little text, suspicious characters, and overlapping text layers.

Understand text layers →

02 / STRUCTURE

The right words.
In the wrong order.

Spot possible column ambiguity and tables whose row or header relationships need preservation.

Understand extraction risks →

03 / CONTEXT

Information beyond
the main text.

See image coverage, annotations, form fields, attachments, and page geometry worth inspecting.

Explore all ten checks →

A SMALL CHECK. A CLEARER NEXT STEP.

Inspect. Understand.
Then decide.

Keep your existing ingestion tools. Preflight helps you see where they may need extra attention.

How the checks work →

01
Choose a source PDF
Upload one file. The server inspects its text and geometry without rewriting the original.
02
Read the evidence
Findings show affected pages and measured signals, with an explanation of why each may matter.
03
Review the right pages
Consider OCR, visual processing, or structure-aware extraction where the evidence points.

FIELD NOTES