From document to a requirements set

Requirements engineering starts with reading. A lot of reading. A 200-page specification, contract annexes, EMVI documents, emails with supplementary agreements — all of these contain requirements, assumptions and obligations you need to understand before you put a single line of design on paper.

The problem is not the volume. The problem is that it is manual work: highlighting, retyping, organising in Excel, and hoping nothing gets missed. And whoever does it determines what comes out.

The DRE takes a different approach.

What the DRE does

The Document Requirements Extractor scans your documents structurally and automatically extracts requirements, assumptions and obligations. Each extracted element receives a unique ID, a source reference, the surrounding context and a classification: is it a genuine requirement, an assumption, an obligation, or not a requirement at all?

The output is a structured table — ready to import into DOORS, Relatics or your own management system. Not as a replacement for those systems, but as the step that comes before them.

Two ways of working: fully automatic, where the DRE runs the document through two passes and processes all relevant segments in parallel, or via human-in-the-loop mode, where you open the PDF yourself, select text segments and choose per selection whether to extract verbatim or rewrite to INCOSE standard.

Verbatim or rewritten to INCOSE

That distinction matters. Sometimes you want to preserve the exact contract wording as a requirement — for traceability and legal certainty. Sometimes you want to convert a vague formulation into a clear, verifiable requirement conforming to the English "shall" standard or the Dutch "dient" formulation.

The DRE does both. You decide per requirement which form you want.

What you get back

Each extracted requirement contains:

• a unique ID (e.g. DRE-0042)
• the source reference (document and paragraph)
• the surrounding text for interpretation
• a short concept-level title
• the requirement text — verbatim or rewritten
• the system of interest and actor
• a classification: requirement, assumption, obligation or not a requirement

You export the output directly to Excel. From there it goes into your requirements management system.

Have a visit

You won't regret it

DRE

You remain the reviewer, not the copyist

The value of the DRE is not only speed. It is in what no longer gets missed: implicit requirements that go unnoticed, formulations interpreted differently by different people, missing source references.

The DRE takes over the extraction work. You review the output, adjust where needed and make the final decisions. The systems engineer remains in control — not of the copy-paste work, but of the content.

Part of a broader workflow

The DRE is the first step in the Basewise SE workflow. Its output feeds directly into the RQA for quality analysis against INCOSE rules, and the REF for linking verification evidence to requirements. From extraction to quality check to verification — as a connected process, not a series of isolated actions.

Curious what the DRE does with your own documents? Book a demo and we will show you.