Knowledge, evidence and reporting
A reference library that keeps analyses grounded in sources
Store the standards, regulations, procedures and papers your project relies on, search inside them, attach them to analysis records as evidence, and let AI answer from your own documents with citations.
Overview
Engineering decisions need to point back to their sources: the clause of a standard, the page of a design report, the result of a test. The Reference Library is the project's curated store of those documents, held as your own copies.
Each reference has metadata, an abstract, documents, citations, notes and a list of where it is used. Documents can be uploaded and indexed or stored only, duplicate files are detected by content, and collections can be shared with other projects.
Indexed documents are searchable down to page ranges, and hybrid keyword and semantic search grounds the AI in your modules. Citations the AI claims are checked against what was actually retrieved, and invented ones are dropped.
Capabilities
Reference records
Metadata, abstract, documents, citations, notes, where-used list and an automation audit.
Collections and sharing
Collections with subcollections, moving and duplicating without copying references, and sharing with other projects.
Upload and index
Upload and index documents for search, or store them only, with indexing status shown.
Duplicate detection
Duplicate files are detected by content wherever they are added.
Search inside documents
Search titles and document text with page ranges and keyboard navigation.
Evidence with roles
Attach references to FMEA records, hazards, requirements and verification results as source, rationale, evidence, assumption, requirement basis or verification evidence.
AI metadata auto-fill
Let AI fill in reference metadata from the document, with a report you can review.
Feeds other modules
Terminology harvest, report reference lists and asset document extraction use library documents.
AI assistance
- Ground AI generation across modules in your documents with hybrid keyword and semantic search.
- Check every citation the AI claims against what was retrieved, and drop invented ones.
- Fill in reference metadata from an uploaded document.
AI proposals are drafts: you review and accept them before they enter your model.
Who uses it
- Engineers who must cite standards and evidence
- Safety assessors and auditors
- Document controllers and technical librarians
- Teams that want AI to answer from their own documents
Works with
- Requirements
References are cited as requirement basis and verification evidence.
- FMEA
FMEA records cite their sources.
- Hazard Management
Hazards cite the documents behind them.
- Terminology
Terminology harvest finds terms in library documents.
- Reports
Reports build reference lists from the library.
- Asset Management
Asset extraction can read library documents.
Frequently asked questions
- Does ModelGraphX include copies of standards?
- No. The library holds your organisation's own copies of standards and documents; no standards content is bundled.
- How does the library improve AI answers?
- Indexed documents are searched with hybrid keyword and semantic search when AI drafts content, and the draft cites the passages it used. Citations that do not match what was retrieved are dropped.
- Can references be shared between projects?
- Yes. Collections can be shared with other projects, and duplicating a collection shares references instead of copying them.
Try it on a real project
Explorer is free, and new accounts get a 30-day trial of the paid capabilities.