Clariti keeps staff review between CivCheck and a plan-review finding

Preserve the AI interpretation as a proposed finding

Clariti's official page explicitly places staff review and approval after each CivCheck AI interpretation. The direct answer is that the interpretation should enter the record as a proposed finding with visible status, not as an agency conclusion. Plans can be incomplete, ambiguous, revised, incorrectly classified, or governed by local amendments and project facts that an automated check does not possess. The reviewer needs to see what the system read, which rule it invoked, and where uncertainty remains.

A proposed-finding record should preserve the submission and sheet version, detected element, cited code text, edition, local amendment set, discipline, location in the plan, explanation, confidence or limitation, creation time, and model or rule version. It should also state whether the issue is missing information, a possible conflict, or a confirmed deficiency. Applicants and staff should not have to infer that distinction from color, rank, or alert severity.

Make code applicability a reviewer decision

A code provision can be current in a content library but inapplicable to the case. Jurisdictions adopt model codes on different dates, add amendments, preserve transition rules, and interpret occupancy, construction type, existing conditions, alternates, phased work, and special approvals through local authority. The case may also depend on zoning, fire, accessibility, utilities, environmental review, historic preservation, or another discipline outside one automated check.

The review screen should expose the code edition and local overlay used, effective date, project attributes that triggered the rule, extracted plan evidence, and any missing input. The authorized reviewer should be able to confirm, revise, dismiss, defer, or request information while recording a reason. Changing the code set or project classification should invalidate or rerun affected interpretations without overwriting the prior record that staff and the applicant saw.

Trace staff disposition into the official correction cycle

Staff approval of an interpretation is still not permit issuance. The retained chain should connect the proposed finding, reviewer action, official correction text, plan location, applicant notice, response, revised sheet, resubmission number, follow-up review, final disposition, and any supervisor or appeals path. Each cycle needs stable identifiers so the agency can distinguish a corrected issue from a rewritten or newly introduced one.

Metrics should keep machine suggestions separate from official findings. Track how many interpretations were reviewed, confirmed, changed, dismissed, deferred, or left unresolved; time to review; reasons for change; code-version errors; and applicant response outcomes. Do not label a suggestion as a violation, a dismissal as model error without analysis, or reviewer agreement as proof of legal correctness. Sampling across permit types and reviewers can reveal where automation needs narrower scope or stronger evidence.

Keep Clariti claims inside the official record

The registered Clariti page establishes current provider positioning for configurable permitting products and CivCheck guided AI plan review. It states that every AI interpretation requires staff review and approval, that AI supports checks, and that code updates can be made. It does not establish interpretation accuracy, complete legal coverage, adoption of a code by a particular jurisdiction, staff authority, a valid correction notice, applicant agreement, or permit approval.

Civic Permit Review reviewed the registered source on August 18, 2026 and did not operate a government deployment. Agencies should verify current supported documents, code sources, local amendments, model and rule versioning, evidence display, reviewer roles, approval states, applicant communications, resubmission lineage, accessibility, records retention, appeals, audit exports, integrations, and service dependencies with representative cases.

Enterprise buyer test

Translate this change into the exact population, record type, workflow stage, decision owner, effective date, and evidence that could be affected. Ask current or prospective providers to demonstrate the named workflow with representative data and an exception—not a polished feature tour. Record what official documentation establishes, what a provider states, what the team observes, and what remains unresolved.

A defensible review also identifies the dependency outside the product. Authority interpretation, policy configuration, data quality, integrations, human judgment, approval rights, release governance, training, and retained evidence may remain customer or service responsibilities. The evaluation should preserve those boundaries instead of treating a technology claim as the complete operating model.

What we will watch next

Civic Permit Review will watch the named source and affected market records for later evidence that changes status, scope, availability, implementation timing, workflow consequence, or the limits of the initial report. A later announcement does not silently overwrite this dated account; the change ledger preserves the sequence.