Model licensing

1.Why this page exists

DocxIntel installs its complete AI engine — including our own models — inside your infrastructure. That is the point of the architecture: the models run on your hardware, so no document has to leave your network. It also means a copy of our models sits on your storage, and you should know exactly what you may do with it before the deployment goes into production.

Buyers usually ask three things. Can we use it commercially, in production, at any volume? Does it keep working if our connectivity is cut? And what are we not allowed to do with the models on our storage? This page answers all three at the level of principle. Your executed agreement is the authoritative text.

2.What ships in the deployment bundle

A deployment bundle contains everything needed to run DocxIntel without reaching the internet. It holds four kinds of material, and keeping them distinct keeps the licensing simple.

DocxIntel models

The proprietary models that power the engine: DocxIntel OCR, DocxIntel Embed, DocxIntel Rerank and DocxIntel Reason, together with the layout, table-structure and classification models. They are developed and owned by BizfyLabs FZC LLC and licensed to you under your DocxIntel licence.

DocxIntel software

The parsing, document intelligence, workflow, agent, search and audit engines, the APIs, the administrative and reviewer consoles, and the deployment tooling. Licensed to you under the same DocxIntel licence.

Open-source dependencies

Inference runtimes, base container images, libraries and system packages, each under its own open-source licence and listed in the notice files of every release.

Manifests and provenance artefacts

The component manifest, the notice and licence files, a software bill of materials, cryptographic digests for every image and model file, and the version pins that let you reproduce exactly what you installed.

3.What you may do with the models

• Run them in production, for your own internal business purposes, across the deployments recorded in your order form.
• Run them fully air-gapped, with no outbound connectivity and no licence check that requires reaching us.
• Process unlimited document volume. There is no per-page, per-document or per-token charge.
• Reprocess your entire archive as often as you like — after a model update, a schema change or a threshold adjustment — at no extra licence cost.
• Keep, index and use every output produced from your documents inside your own systems, without restriction from us.
• Run them on the infrastructure you choose: bare metal, your own data centre, your private cloud tenancy or a sovereign cloud region.
• Copy the bundle for backup, disaster recovery and staging for as long as your licence runs.
• Have your own security team inspect the artefacts, verify the digests and scan the images before installation.

4.What you may not do

The models are our intellectual property running on your hardware. The limits below protect them; they are set out in full in the terms and conditions and the acceptable use policy.

• Do not redistribute the models, publish them, upload them to a model hub, sell or sublicense them, or pass them to a third party outside your licence.
• Do not extract the models for use in a separate system or product, and do not distil, clone or attempt to reconstruct them.
• Do not reverse engineer the models or the software to derive architecture, parameters or training data, except where mandatory law expressly permits it.
• Do not use the deployment to provide a document-processing service to third parties unless a written reseller or service-provider addendum permits it.
• Do not retain any part of the bundle after your licence ends; the software and the models must then be removed.

5.Your data and our models

• Your documents, extracted values and feedback never leave your deployment and are never used to train, tune or improve any model we ship — for you or anyone else.
• Any tuning for your document types happens inside your environment and stays there.
• Model improvements reach you as signed, versioned offline packages that you validate and promote on your own schedule, with the previous version kept for rollback.

6.SBOM and provenance artefacts

• A software bill of materials in a standard machine-readable format, covering container images, system packages, language dependencies and model artefacts.
• Cryptographic digests for every container image and every model file, so an installed deployment can be matched against the published release.
• Signatures on release artefacts, verifiable offline before installation in an air-gapped environment.
• Notice files listing every open-source component, its version and its licence.

The manifest for a specific release is available on request at https://bizfylabs.com/contact-us, and is part of every Proof of Value pack so your legal and security teams can review it during the engagement.

7.The authoritative licence text

This page describes the licensing model to help you evaluate the product. It is not a licence and grants no rights. Where anything here differs from your executed licence agreement or order form, that agreement governs.