Choose what you'd like to import

Drop a Word, PDF, Excel, or Markdown file. Lucora's AI will extract candidate requirements, design specs, tests, and trace links — nothing is committed until you've reviewed each one.

Drop a file or click to browse

Word (.docx) · PDF · Excel (.xlsx) · Markdown · plain text · up to 50 MB

CGM-v2_PRD_RevC.docx

38 pages · 412 KB · uploaded just now

Document parsed. Identified 5 sections matching common SRS structure (Goals, System Requirements, Software Requirements, Design, V&V).

Extraction model

Optional hint to the model

Extract trace links between items Retain source passages for re-run Process without uploading to model (BYO key)

Confirm what we found

We've inferred how this document maps to V-model artifact types. Adjust if needed — anything you change here will guide extraction.

Detected sections → artifact types

${''/* populated by JS */}

Sample passage

2. System Requirements
From the goals above, the primary system requirements are derived. The sensor shall report filtered glucose readings at a 5-minute cadence; battery life at expected usage shall meet or exceed 14 days. Alarm thresholds are fixed in the initial product release: the device shall alarm on BG < 70 mg/dL, and shall alarm on BG > 250 mg/dL…

Highlights show what we'd extract under the current mapping. If the AI is misreading a section, change the mapping on the left.

How should we handle identifiers?

Your document already uses identifiers like UR-01, REQ-12, SW-101. Choose what becomes the canonical Lucora ID — and what's preserved as a reference. This is one of the most consequential choices for regulated workflows.

Once items are baselined and signed, identifiers are difficult to change. Renumbering an audited artifact is treated as a controlled change and requires Part 11 justification. Choose Preserve if your existing protocols, emails, and external documents already reference these IDs.

Recommended · default

Preserve source IDs

Source identifier becomes the Lucora display_id. No prefix is applied. Existing docs, test protocols, and external references continue to resolve unchanged.

Best for: mature projects already in active use; anything that's been previously reviewed by an auditor.

UR-01 → UR-01 unchanged

Assign fresh Lucora IDs

New display_id follows the project's prefix scheme. Source identifier is preserved as legacy_id — searchable and visible as an optional column.

Best for: new projects still in flux; consolidating multiple source documents with overlapping ID schemes.

UR-01 → UN-0001 + legacy_id

Hybrid

Preserve IDs for items already at approved status in the source. Assign fresh IDs for drafts. The "right answer" for most regulated teams.

Best for: mid-life projects with a mix of approved and in-flight items.

UR-01 → UR-01 approved

Live preview

First 7 items as they would appear in Lucora under the selected mode

Project prefix scheme

Type	Source ID		Lucora display_id	legacy_id	Status	Title

Resolve identifier conflicts

Three conflicts detected. Resolve each before proceeding — Lucora won't import anything until display_id is unique within the project.

Within-import collision

Two extracted candidates both claim source ID REQ-42. Common when a PRD revision introduced a renumbering.

SRREQ-42p.8 §1

PHI shall be encrypted at rest

SRREQ-42p.8 §3

App shall authenticate via OAuth 2.0

Conflict with existing project item

Source ID T-U-06 matches an existing item already in this project: UT-0104 · test_watchdog_reset. Likely the same artifact being re-imported.

Prefix scheme mismatch

Source uses UR- / REQ- / SW- / DSN- / T- / VV- / CLN-. Project prefix scheme is currently UN- / SR- / SS- / DS- / UT- / IT- / VT- / VAL-. Under the chosen Preserve mode, source IDs override.

Ready to extract

Lucora will run extraction with the settings below. You'll review every candidate item and link before anything is committed.

Estimated items

~42

across 8 V-model types

Estimated links

~29

trace edges between items

Conflicts

3

all resolved

Est. cost

$0.18

~12 sec runtime

Settings summary

Source document	CGM-v2_PRD_RevC.docx · 38 pages
Target project	CGM-v2 · acme-biotech
Extraction model	Claude Sonnet 4.6
Identifier mode	Preserve source IDs · source identifier becomes display_id
Project prefix scheme	Adopt source scheme (UR- / REQ- / SW- / DSN- / T- / VV- / CLN-)
Trace links	Extract proposed links between items
Source passages	Retained for re-run with hint
Conflicts resolved	3 of 3 (1 collision · 1 existing-item merge · 1 prefix decision)

Nothing in your project will be modified until you've reviewed each extracted candidate on the next screen and confirmed commit. Source-document attribution is retained for every accepted item — required for downstream Part 11 traceability.