InspectMind AI logo InspectMind
Research library

Production snapshot ·

InspectMind Plan Review Metrics & Methodology

Verified 2026-09-16: 1,435 completed customer project records, 1,490 checks, and 155,970 retained potential findings. Definitions and exclusions.

These counts describe the retained review records in InspectMind’s production system. They measure review activity and potential findings; they do not establish accuracy, code compliance, or realized savings.

Completed project records
1,435
Distinct project records with a completed check. Separate records may refer to the same building; this is not a count of unique buildings.
Completed checks
1,490
Distinct check records with at least one completed production run. Retried runs do not increase this count.
Retained potential findings
155,970
Potential findings linked to the counted checks. Duplicate-group members, user-dismissed findings, wrong finds, AI-dismissed findings, insufficient-reference findings, and previously suppressed findings are excluded. Remaining findings are not all independently validated.
Recorded pages across checks
294,986
Sum of positive recorded page counts on completed checks. Missing counts are excluded; repeated checks can include the same pages. This is a minimum of recorded page submissions, not unique sheets or proof of complete page coverage.

What is excluded?

Deleted projects, InspectMind-owned projects, super-admin-owned projects, and projects named with a test, demo, sample, or untitled prefix are excluded.

Why do these numbers differ from earlier claims?

Snapshot of retained production records, not an all-time historical ledger or an industry-wide defect or accuracy rate. Definitions differ from earlier marketing totals. Earlier website figures used broader or historical populations. This snapshot consistently links findings to completed checks and excludes rejected findings. It should not be interpreted as a growth comparison with those older totals.

How complete is the page count?

542 completed checks have no positive recorded page count and contribute no pages to this total. Consequently, we present pages as a recorded minimum. Repeated checks can count the same document pages again.

How should findings be interpreted?

A retained finding is a potential discrepancy for professional review. A “good catch” label records feedback within the product; it is not an independent accuracy assessment. Counts vary with document scope, project stage, and review settings. We do not infer an accuracy percentage or dollars saved from finding counts.

Download and cite

Suggested citation: InspectMind AI, “Plan Review Metrics & Methodology,” production snapshot dated 2026-09-16. Include the snapshot date and definitions when citing a count.

Download aggregate JSON

The download contains aggregate counts and definitions. It contains no customer identities or private project records.

Inside the review

See the conflicts before you upload.

Real document discrepancies, the reason they need attention, and a concrete next step for the design team.

Open a sample report

Commercial · Anonymized review

Fire-resistive curtainwall framing disagrees

Compare architectural assemblies, shutdown planning, and access-control sequences before procuring laboratory building systems.

Read the findings

Commercial · Anonymized review

Floor elevations differ across disciplines

Check that structural load criteria and building elevations agree before framing and truss design proceed.

Read the findings

Education · Anonymized review

Smoke-detector scope differs

Compare school equipment schedules with controls specifications and the document index before coordinating MEP installation.

Read the findings

Review the evidence, check the price, then try your own set. Eligible work emails receive a $100 first-check credit.

Browse all project studies