Skip to content

Enterprise

Verify files on your own infrastructure, or ours.

Fileverity checks images, text and documents for AI generation and tampering, and issues reports that others can verify. Enterprise adds volume pricing, on-premises deployment and integration with your claims, onboarding or editorial systems.

  1. 01Deployment
  2. 02Checks
  3. 03Integration
  4. 04Security and governance
  5. 05Process
  6. 06Industries

Deployment

Hosted service or your own servers.

Choose where the analysis runs. Both options use the detection models Fileverity develops in-house: VerEye for images and Veread for text.

Hosted service

Use the web application, the Console and the API from one workspace, with one shared credit balance. Enterprise services are billed as stated in your agreement.

  • Image, document and text analysis runs on Entogo-operated servers in Toronto, Canada. The web application runs in the United States.
  • Uploaded files are deleted after analysis. Submitted text is not stored.
  • A verification record is kept for each report: its ID, time of issue, the file’s SHA-256 hash and the findings.

On-premises deployment

Run Fileverity on your own servers, fully offline, with no external API. Sensitive and regulated files never leave your infrastructure.

  • No call to any external service, and no exposure to the public internet.
  • Analysis runs on VerEye and Veread, developed in-house.
  • Offered to Enterprise customers on request. We scope the deployment with you.

Checks

Evidence for images, text and documents.

Each check supports a decision made by a person. Reports state what they cannot show.

  • Images

    AI image detection assesses whether a whole image was generated by AI, with the checks, heat map and closest generator families behind the result. The image verification report adds local edits such as splicing, copy-move and generative fill, and reads Content Credentials (C2PA), metadata, AI labels and watermarks. The report gives a risk level from L0 to L5 and a recommended action, as a signed PDF and as JSON.

  • Text

    AI text detection assesses whether text is human-written, AI-written or mixed, and highlights the passages that show signs of AI writing. It works in seven languages and accepts up to 100,000 characters per check. The result identifies passages for review. It does not identify the author.

  • Documents

    One document report covers the text, up to 15 embedded photos and scanned pages, quoted changes between saved versions, editing software and dates, hidden content, scanned amounts and dates, and PDF signatures. Files above 50 MB or 100 pages are handled under Enterprise. Each report carries a report ID, the file’s SHA-256 fingerprint and, when validated, an RFC 3161 trusted timestamp and a PDF seal.

Integration

Fit the checks into your existing process.

Reports come as JSON for your systems and as print-ready PDFs for people, in seven languages.

  • API

    One API covers image, text and document analysis. Keys are bearer tokens created and revoked in the Console, with a monthly spending limit per key. You can wait for a report in the same request or receive it by signed webhook, and the API is described by an OpenAPI 3.1 specification. Test keys return fixed sample results and are never charged.

  • Batch processing

    Bulk checks are available to workspaces that have purchased credits. For larger files and higher volumes, contact us about Enterprise options.

  • Claims, onboarding and editorial systems

    Store the report ID beside your own case or customer record, and print your case reference on the PDF. Anyone you share the ID with can confirm the report on the public verification page without an account.

  • Custom development

    Enterprise agreements can cover detectors adapted to your file types, risk policies, report formats and system integrations for specialist sectors. We define the scope with you and evaluate it on your own files.

Security and governance

What your security and procurement teams will ask for.

This section summarizes our published trust, privacy and data processing pages. A written agreement with Entogo takes precedence where it differs.

Data handling and retention
Uploaded files are deleted after analysis, and customer files are not used to train or tune detection models. Temporary evidence is deleted from the server after delivery is acknowledged, and undelivered results expire. Verification records are kept for as long as reports may need to be verified, or as agreed with you or required by law.
Security practices
Traffic uses TLS with HSTS on every subdomain, and stored data is encrypted at rest by our infrastructure providers. The analysis engines are reachable only through an authenticated, firewalled tunnel that accepts only Fileverity’s application. Sign-in uses one-time email codes or Google, and Fileverity stores no passwords.
DPA and vendor reviews
A data processing agreement is included with Enterprise. Entogo acts as processor and processes personal data in submitted files only on your documented instructions. The current list of subprocessors is published on the trust page, and customers are informed of changes. Our security and risk team handles security reviews and vendor questionnaires directly, and supports model-risk reviews, for example under OSFI Guideline E-23 or the NAIC AI Model Bulletin, with documentation on request.
Verifiable reports and stated limits
Each report names the exact file through its SHA-256 hash. A trusted timestamp or PDF seal appears only when its validation succeeds. Findings are evidence for a human decision, not a certification, and “no anomaly found” never means “proven authentic”. Fileverity is not intended for denying a claim or rejecting a person on the result alone.

Process

How an Enterprise engagement runs.

The steps are the same whether you choose the hosted service or an on-premises deployment. Timing depends on your case.

  1. 01 / 04

    Describe your case

    Tell us which files your teams review, the approximate volume and the systems involved.

  2. 02 / 04

    Evaluate on your own files

    We define the scope with you and test the checks on files from your own workflow.

  3. 03 / 04

    Agreement and documents

    We agree on the commercial terms and the deployment option, and provide the DPA and the documentation your vendor and security reviews need.

  4. 04 / 04

    Rollout

    We deploy hosted access or the on-premises installation and connect the checks to your systems, with onboarding and priority support.

Industries

Evidence for professional decisions.

  • Insurance claims

    Check claim photos and supporting documents before settlement for image alterations, document revisions, scanned amounts and dates, signatures and AI-generated content. Adjusters and SIU teams see the source evidence behind each finding.

  • Financial services

    Examine onboarding and lending files for image alterations, document inconsistencies, available revisions and signature findings. Review scanned fields and AI-written passages alongside your issuer and account checks.

  • Newsrooms and fact-checking

    Review submitted images and contributed text before publication, and examine available Content Credentials to help establish provenance.

  • Legal, investigations and compliance

    Record the findings of an image or document analysis and link them to the exact file through its SHA-256 fingerprint. Counsel and auditors can inspect the report’s verification record independently.

Questions from security and procurement reviewers

Where are files processed, and how long are they kept?

Image, document and text analysis runs on Entogo-operated servers in Toronto, Canada. The web application runs in the United States, and uploads are held temporarily in Montréal, Canada until analysis completes. Uploaded files are deleted after analysis, and submitted text is not stored. We keep a verification record for each report for as long as reports may need to be verified, or as agreed with you or required by law.

Can Fileverity run inside our own infrastructure?

Yes. Enterprise customers can run Fileverity on their own servers, fully offline and with no external API, so files never leave their infrastructure. This is offered on request and scoped with you through our team.

Do you sign a data processing agreement, and who are your subprocessors?

Yes. A data processing agreement is included with Enterprise. The current list of subprocessors, with the purpose and location of each, is published on the trust page, and customers are informed of changes.

Does Fileverity hold ISO/IEC 27001 or another security certification?

The frameworks we design against, including NIST Cybersecurity Framework 2.0, ISO/IEC 27001 and the NIST AI Risk Management Framework, are not certifications of Fileverity. Members of our security and risk team hold professional certifications, including ISACA’s CISA, CISM, CRISC and CDPSE. For your own assessment, our security and risk team answers vendor questionnaires directly.

Can a report be relied on as proof that a file is authentic or altered?

No. A report is evidence for a human decision, and it states what it cannot show. A third party can look up the report ID on the public verification page and match the exact file by its SHA-256 hash. The trusted timestamp shows when the report data existed, not that the findings are correct.

How is Enterprise priced?

Enterprise uses volume pricing by agreement, and services are billed as stated in your agreement. Pricing is discussed case by case, based on your file types, volume and deployment. Contact us to start.

Tell us which files your teams review.

Send us the file types, the approximate volume and the systems involved, and we will discuss the right deployment for your organization.