The professional document pipeline that runs on your machine

Professional document work is sequential, and the stakes rise at every stage. A contract has to be assembled in the right order before annexes are numbered; a filing has to be split into exhibits before each one is stamped; a regulated dataset has to be redacted before it ever reaches a recipient. Xi Tools is organised around that reality instead of scattering fifty icons across a homepage. Each stage below maps to dedicated tools, and because the page-level operations execute locally, you can run a four-step sequence — compress a scanned bundle, split out the privileged pages, add page numbers, watermark the draft — without a single confidential file leaving the machine. That continuity is the whole point: one controlled pipeline rather than four separate uploads to four separate services.

Most PDF sites treat every operation as an isolated one-off, which is fine until you have a 200-page bundle that needs to come apart, get reordered, be stamped, and go back out as separate exhibits. Xi Tools is built for that chain. The tools below are listed in the order a real document moves through them: you split the source into its working pieces, reorder and merge them into the sequence a reviewer expects, stamp page numbers and a watermark so every reference is unambiguous and every draft is marked, then export — as a single locked PDF or as page images for a deck. Each step hands a clean file to the next, and because the page-level work runs in the tab, a confidential bundle never has to be uploaded between stages to keep the pipeline moving.

Conversion is where document work usually breaks down, because the format your authoring tool produces is rarely the format the next desk accepts. This matrix maps each source format to its supported targets and, just as importantly, to the moment you actually need it: locking a Word draft into a fixed-layout PDF before it goes external, rebuilding an editable DOCX out of a received PDF, or moving a dataset between CSV, JSON and YAML as it passes from spreadsheet to API to deployment config. Read down the Source column to find the file you have, then across to the target a recipient, a portal or a parser expects.

The catalogue looks broad, but it resolves into three coherent disciplines that a document professional moves between all day. Page-level PDF editing is everything you do to a bundle without changing what kind of file it is — the assembly, splitting and stamping that produce a clean filing. Format conversion is the bridge between the format you authored in and the one a recipient demands, which is where most friction actually lives. Data-format wrangling handles the structured side: spreadsheets and configs that have to move between CSV, JSON, YAML and Excel as they pass from analyst to engineer. Knowing which family a task belongs to is usually half the work of finding the right tool.

Document work carries a compliance burden that ordinary file handling does not. A signed contract, a payslip run, a patient record, an unredacted discovery bundle or an unreleased financial filing is precisely the category of material that data-protection rules, client engagement terms and internal handling policies say must not be casually uploaded to a third party. That obligation is the design constraint Xi Tools is built around: the great majority of these tools run client-side, so the operation happens inside the browser tab on your own hardware and the regulated document is never transmitted to a server we or anyone else controls.

The consequences matter in a compliance review. Data residency: for client-side tools, the file stays on the machine and inside the network boundary it started in — there is no copy sitting in an external upload bucket, no third-party processor to add to a data-flow map, and nothing to issue a deletion request for afterwards. Auditability: because nothing is transmitted for those operations, there is no remote log of your document's contents to reconcile, and the chain of custody stays where your policy already covers it. Throughput without a queue: the work runs at the speed of your own hardware rather than waiting on a shared remote job, which also means it keeps working on a restricted or air-gapped connection once the page has loaded.

There are honest limits, and a compliance-minded team should know them up front. A few conversions — notably high-fidelity DOCX to PDF and PDF to DOCX, where rendering depends on heavyweight layout engines — are genuinely hard to perform faithfully in a browser alone, and the right tool for those formats may do part of its work server-side. We label that on each tool's page rather than blur it. The rule of thumb: the page-level PDF operations (merge, split, compress, rotate, page numbers, watermark) and the lightweight text and data conversions (Markdown, HTML, CSV, JSON, YAML) execute locally, and that is the bulk of the catalogue. In every case there is no account, no service watermark on your output, and no usage ceiling — the controls stay with you, not with us.

Most document problems are procedural, not technical — the hard part is knowing which tool to reach for and in what order, not operating any one of them. These guides walk through the sequences that come up again and again in legal, finance and operations work: assembling a clean contract pack, splitting a bundle into routable exhibits, getting a scan-heavy filing under a portal limit, and choosing sensibly between CSV, JSON and YAML for a dataset. Each one is written for the person doing the job rather than the engineer who built the tool, with the compliance considerations called out where they matter — so you can follow the steps once and then run them from memory.

A document task is rarely one operation — it is a short chain of them, and the value is in running that chain cleanly from intake to a filed result. The jobs below are the ones that fill a working day on a contracts, filing or operations desk, and each is described as the actual sequence of Xi Tools you would reach for, in order. Notice the common thread: because the page-level steps execute on your own machine, a confidential bundle moves through assembly, stamping and compression without being uploaded between stages — so a multi-tool workflow stays inside the same handling boundary your policy already covers, start to finish.