Use cases
Mapped to features that exist
Six kinds of work, each linked to capabilities the shipped build performs — with the tests that cover them.
Office workers
Anyone who touches contracts, letters and forms daily
- Fix a typo in a signed-off PDF→ Edit existing text (font preserved)
- Add a signature and today's date→ Signatures: draw, upload or type
- Combine a packet before sending→ Merge
- Lay 4 pages per sheet for a handout→ Print — N-up with true paper size
In practice
Edit → Save → Print, without ever exporting to another tool.
Students
Reading scans, handouts and paper chapters
- Search a scanned chapter for a quote→ Searchable OCR + full-text search
- Pull passages into an essay draft→ OCR editable, then copy or save .txt
- Turn phone photos of notes into Word→ Image to Word
- Shrink a 200-page reader before sharing→ Compress with target size
In practice
Paper becomes searchable text you can quote exactly.
Accountants & finance teams
Statements, invoices and ledgers
- Get a statement's table into a spreadsheet→ PDF to Excel (.xlsx, cells read as numbers)
- Find every amount across the library→ Full-text search with snippets
- Compress a quarter of scans to a size limit→ Target-size compression (7 measured passes)
- Keep working on a plane→ 100% offline — no connection needed
In practice
Numbers arrive as numbers, and the app proves it opened them.
Compliance, legal & HR
Documents that must not leak
- Remove a person's details permanently→ Redaction — characters deleted, not painted over
- Lock a file before it circulates→ Password protection (AES-256)
- See what changed between revisions→ Compare — word diff + page pictures in one report
- Never invalidate a digital signature by accident→ Digitally signed PDFs are refused, with the reason
In practice
Everything stays on the machine — nothing to subpoena from a cloud.
Document processing teams
Recurring jobs over many files
- Watermark / rotate / number 200 files→ Batch operations (7 jobs, CSV report)
- Know exactly which files failed→ Per-file outcomes + retry-only-failed
- Stop a run that went wrong→ Stop honoured between documents and between OCR pages
- Standardise A4 / Letter output→ Batch resize (A3/A4/A5/Letter/Legal)
In practice
One pass, one CSV, honest per-file results.
IT administrators
Deploying to managed Windows PCs
- Install without admin rights→ Per-user Inno Setup installer (no elevation)
- Deploy a self-contained bundle→ Professional ZIP with its own Python runtime
- Diagnose a user's machine remotely→ Health check + exportable diagnostic report
- Enable sign-in for shared machines→ Optional multi-user: roles, PBKDF2, rate limiting
In practice
No Python, no internet, no package manager on the target PC.
Same tool, every document type
Nothing above is a mock-up workflow: each line names a shipped capability with a regression test behind it.