eriCoord
  • Documentation
  • Pricing
  • FAQ
eriCoord

VeriCord identifies candidate conflicts between drawings and specifications so a qualified engineer can review them before construction.

© Copyright 2026 VeriCord. All Rights Reserved.

About
  • Contact
Product
  • Documentation
Legal
  • Terms
  • Privacy
  • Security
  • Acceptable Use
  • Subprocessors
  • Accessibility
  • Documentation
  • Pricing
  • FAQ
  • AI and Data Governance

    How VeriCord governs automated analysis, traceability, and customer data.

    Last updated: October 6, 2026.

    VeriCord uses automated analysis, including artificial intelligence where appropriate, to support document review and project questions. This statement describes how that processing is governed. It should be read with the privacy notice, the security page, and the professional judgment disclaimer.

    Human oversight

    VeriCord does not replace qualified professional judgment. Candidate findings and Ask VeriCord responses are for review by qualified users. Final engineering, architectural, construction, contractual, regulatory, and safety decisions remain with the responsible professionals and organizations.

    Traceability

    Where the product can do so, substantive findings and project-intelligence responses are linked to the project documents, specification sections, drawing sheets, review findings, or external references that support them. The basis of a result is meant to be reviewable. VeriCord does not ask you to accept an unsupported conclusion as a determination.

    Data governance

    • Customer data is processed only for the purposes in the terms of use, the privacy notice, and any applicable customer agreement.
    • Access to customer workspaces is limited to authorized users and to operational personnel with a business need.
    • Customer Content is not used to train third-party foundation models or shared models for other customers unless the customer expressly agrees in writing.
    • Subprocessors and model providers are those listed on the subprocessors page, under the controls described there and on the security page.

    Models and risk

    A review may combine document extraction, retrieval, rules, and language or vision models. VeriCord may change underlying technologies over time. Public claims describe outcomes and controls. They do not disclose proprietary routing logic, prompts, thresholds, or model orchestration.

    Material changes are tested before production release. Known failure modes are treated as product defects to be reviewed, not as determinations a user should accept without inspection.

    Accuracy and candidate findings

    Automated systems can produce incomplete or incorrect results. VeriCord treats outputs as candidate findings and decision-support information. The product keeps source inspection and a place to record a reviewer decision so a qualified user can accept, reject, or correct a finding.

    External standards

    VeriCord may help identify potentially relevant codes, standards, agency criteria, and discipline references. An external standard is not treated as governing a project unless the available context supports that reading or you confirm it. You must still check version, jurisdiction, and contractual applicability, as the disclaimer describes.

    Change management

    Material changes to review behavior are tested before they reach production. Changes that affect what customers see are recorded through the release process appropriate to their impact.

    Security and confidentiality

    AI-enabled processing is subject to the same security, confidentiality, access-control, retention, and incident-handling expectations as other VeriCord processing. Customer project information is not sent to public consumer AI services outside the approved production integrations on the subprocessors page.

    Responsible use

    Use VeriCord for lawful professional work. Do not use it to bypass required professional review, to misrepresent authorship or certification, to violate client confidentiality, or to claim code, safety, or contractual compliance that the documents and a qualified reviewer do not support. The acceptable use rules apply.

    Contact

    Questions about these practices can be sent through the contact page.