VANTAGE

VANTAGE is Struxen's automated quality control engine. While STRUX answers a single question, VANTAGE runs a full archetype — a structured multi-step analysis — across your project documents and returns a list of findings with severity levels, source citations, and suggested follow-up actions.

Pricing

VANTAGE is an optional add-on at $250/month per active project. Toggle it on from the pricing page when you sign up, or add it to a project later. The rest of the platform does not require it. See Account & Billing for how it fits the overall pricing model.

When to run VANTAGE

Run VANTAGE at natural project milestones, not continuously. Good trigger points:

• End of design development — coordination check across disciplines
• CD set issued — drawing-to-spec consistency, completeness check
• Bid set received — scope completeness review for subcontractors
• Addendum or revised drawings posted — re-run to surface new issues
• Pre-bid or pre-construction milestone — final QC sweep

A VANTAGE run takes a few minutes and consumes Compute Credits scaled to the number of documents and the archetype depth. Run it when you need a comprehensive answer, not as a constant background process.

Selecting an archetype

An archetype is a pre-built analysis recipe. When you click New Run you choose one. Available archetypes include:

CrewSense

Looks at labor and crew coordination across trades to flag stacking conflicts and crew-density issues.

Each archetype description in the UI lists its inputs (which document types it reads) and its typical run time.

Running VANTAGE

1. Open VANTAGE in the project sidebar.
2. Click New Run.
3. Pick an archetype.
4. Optionally narrow the scope (specific folders, disciplines, or document types). The default scope is the entire project.
5. Review the credit estimate, then click Start Run.

The run progresses through phases: Queued → Retrieving → Analyzing → Synthesizing → Complete. Most runs finish in 2–8 minutes. You can navigate away — runs continue server-side.

Interpreting findings

When the run completes you land on the findings page. Each finding has:

• Title — one-line summary
• Severity — Critical, High, Medium, Low, or Informational
• Discipline tag — which trade(s) the finding affects
• Description — the analysis in 2–4 sentences
• Source citations — every document and page the finding was derived from, with click-through to the original
• Suggested follow-up — concrete next action (e.g. "Issue an RFI to confirm…", "Update spec section to match drawing…")

Severity levels

Severity	Meaning
Critical	Likely to cause significant cost or schedule impact if not addressed before construction. Treat as immediate.
High	Material conflict or omission that warrants a resolution this week.
Medium	Inconsistency or unclear requirement that should be resolved before procurement.
Low	Minor wording or formatting issue. Address opportunistically.
Informational	Observation only — no action required, but worth knowing.

Severity is assigned by the model based on construction impact, not by keyword matching. Two findings with the same root cause can carry different severities depending on which trades and which schedule windows they affect.

Acting on findings

Each finding has three actions on its detail row:

• Convert to RFI — drafts an RFI in the project RFI module with the finding text and citations pre-filled. Edit and send.
• Mark resolved — closes the finding when the underlying issue is fixed. Resolved findings remain in the run history.
• Dismiss — for findings that are not real issues (false positives). Dismissed findings stay in the record so reviewers can see why.

Re-running

When drawings or specs are revised, click Re-run on the original run to apply the same archetype to the updated documents. The new run is versioned — old findings remain accessible.

Troubleshooting

Run stuck in "Analyzing" for >15 minutes. Click Cancel, then start a fresh run. If it stalls again, email support@struxen.io with the run ID.

Findings reference the wrong drawing. Confirm the right document versions are attached. If versions are correct, dismiss the finding with a note — your dismissal feedback is used to tune the archetype.

Credits are higher than expected. Run cost scales with document count and archetype depth. Scope to specific folders for targeted runs.