Skip to content
GearDock

Platform

One System for Medical Catalog Content, Not a Writing Tool

Product data and manufacturer documents go in one end. Reviewed, approved, exportable catalog content comes out the other, with the evidence trail still attached. This page explains how the parts fit together and why they are arranged this way.

Design Principles

Four Decisions That Shape Everything Else

Most of what surprises people about GearDock follows from these four choices. They are deliberate, and they are why it does not behave like a general-purpose writing tool.

  • One Record, Every Stage

    Import, matching, evidence, completion, drafting, review, release, and export all read and write the same product record. Nothing is exported from one tool and imported into the next, so nothing is lost at the handover.

  • The Source Stays Attached to the Claim

    A sentence written from a document passage keeps its link to that passage through editing, rewriting, review, and export. Traceability is part of the record itself, not a report someone assembles afterwards.

  • A Gap Is a Result, Not a Failure

    When nothing supports a value, GearDock says so rather than producing something plausible. An empty field gets chased. A confident figure nobody can check does not.

  • People Make the Decisions That Carry Weight

    Uncertain matches, review, approval, and release are human decisions. There is no setting that lets the system approve or publish on its own.

System Map

Nine Parts, One Shared Record

Grouped by what each part achieves rather than how it is built. They share the same product record underneath, which is what makes this a system rather than a diagram.

Bring data in

  • Product import
  • Document intelligence
  • Product matching

Make it complete

  • Content completion
  • Governed generation
  • Claim Firewall

Prove and distribute

  • Review and release
  • Digital thread
  • Integrations

One governed product record underneath. Evidence attached at intake remains available during review, release, distribution, and later correction.

Layer by Layer

What Each Part of the System Does

Nine layers, each with one job and a page that goes deeper on it.

LayerWhat it doesRead more
IntakeTakes spreadsheets and product records, maps the columns, checks every row, and holds the batch for you to look at before anything is committed.Product import
EvidenceKeeps your source documents in your organization's library, reads their contents, and links them to the products they cover.Document intelligence
IdentityWorks out which real product each row is, using identifiers first and evidence after, recording how confident the match is and sending the uncertain ones to a person.Product matching
CompletenessCompares what is known against what a product of that type normally carries, and marks each value supported, weak, conflicting, or missing.Content completion
WritingDrafts the title, summary, description, and specifications from those facts and the passages behind them, keeping the citations attached.Governed generation
CheckingCompares every statement in the draft against the evidence and the configured rules, and blocks release when a claim has nothing behind it.Claim Firewall
DecisionShows the draft beside its evidence for a person to review, and records approve, revise, or reject with a reason and a name.Review and release
HistoryKeeps every version, source link, decision, release, and delivery for a product as one ordered history that nobody can edit later.Digital thread
DistributionPackages released content as a file or sends it to a connected system, re-checking the release state at the moment it runs.Exports and integrations

What Your Team Touches

The Complicated Part Stays in the Platform

Catalog staff learn four actions: import, confirm a match, review a draft, release it. Everything below that line runs without configuration, which is what lets a small team publish a large catalog without first becoming governance experts.

What your team does

  • Upload
  • Review
  • Approve
  • Export

runs on

What the platform maintains

  • Evidence
  • Matching
  • Confidence
  • Claim controls
  • Versioning
  • Audit
  • Lineage
  • Organization isolation

The Record Underneath

Every Value Knows Where It Came From

This is the shape of a product record in GearDock: a field, its value, and the evidence state behind it. Gaps and conflicts are proper states in their own right, not errors to be cleared.

  • A supported value carries the document and passage it came from.
  • A gap stays empty and visible instead of being filled with a guess.
  • A disagreement between two documents comes to you rather than being settled quietly.
  • That state follows the value through drafting, review, and export.
Product recordAcme Medical AM-4200
AttributeValueEvidence
ManufacturerAcme Medical
Source-backedAM-4200 datasheet, p.1
ModelAM-4200
Source-backedAM-4200 datasheet, p.1
Flow rate range0.1-999 mL/h
Source-backedAM-4200 datasheet, p.4
Weight2.4 kg
Source-backedService manual rev. C, p.12
Battery runtimeNo source
GapNo source supports this
Power input100-240 V AC
Conflict2 sources disagree

Illustrative example using fictional product data.

What GearDock Is Not

It is not a PIM you configure and fill in yourself. It is not a document management system. It is not a general-purpose AI writer, and it is not a pipeline that publishes on its own. It is a workflow that turns product data and manufacturer documents into approved catalog content, with people making the decisions that carry responsibility. If you already run a PIM, GearDock feeds it rather than replaces it.

Private Beta

Bring Us a Slice of Your Catalog

The quickest way to judge GearDock is on your own products. Send a representative spreadsheet and whatever documents you hold for it, and we will run the workflow with you.