ModelNex Product & Browser Extension Privacy Policy

Scope and responsibility

Precept Labs provides ModelNex to organizations that create, run, and improve workflows. This policy describes the product and browser extension. Our separate website Privacy Policy covers visitors to the marketing website.

Your organization determines the workflows, connected applications, user access, deployment location, and service providers configured for its deployment. Customer agreements govern processing on behalf of the organization. Ask your administrator which deployment and processing arrangements apply to you. Questions about this policy can be sent to support@preceptlabs.ai.

Account and workflow information

ModelNex processes account identifiers and access information, such as your name, email address, organization, permissions, and authentication information needed to verify your access. Depending on the features you use, it also processes chat messages, workflow instructions, files you submit, answers and approvals, connected-system results, and run history.

This information can include personal or sensitive business information contained in your organization’s workflows. Only use ModelNex for information your organization has authorized you to process.

Browser recording

When you start a recording, the extension can capture your interactions, entered field values, page URLs and titles, visible page content, notes, timestamps, and screenshots for session replay. It can capture the application’s network requests and responses, including methods, URLs, headers, and bodies, to help reconstruct the workflow. Recording can follow you across eligible application tabs until you stop it or its recording window ends.

These recordings can contain personal information, communications, health information, financial information, or other sensitive data displayed or entered in the application. The extension does not guarantee that sensitive values are removed from page content, screenshots, URLs, or responses. Do not record information you are not authorized to share with your organization’s ModelNex deployment.

The extension does not read your saved browser-history database. URLs and page transitions encountered during an active recording are part of that recording. Installing the extension alone does not start a recording.

Observation and user controls

Separately configured observation can collect semantic workflow events on organization-approved sites for users with observation access. It has a visible indicator and pause control. Observation access is separate from ordinary recording access.

You can stop a recording, pause observation, sign out, or disable or uninstall the extension. Those actions prevent further use as applicable but do not delete recordings or other data already submitted to your organization’s server.

Where information goes

The extension communicates with the ModelNex server configured for your deployment. Recording chunks and replay information may be backed up to that server while recording is in progress, before final submission. Submitted recordings are available to authorized reviewers; recording a workflow does not automatically publish changes to it.

Execution/chat recording uses a short-lived recording grant verified by the configured server. The surface does not pass its identity-provider access token to the extension through the recording bridge. Portal-based access can use your signed-in portal session. Local extension storage holds configuration, access state, recording recovery data, and replay frames needed to operate these features.

Customer-managed deployments can run in the customer’s environment. Hosting location and any transfers to other regions depend on the deployment and connected services your organization selects.

Purposes and recipients

ModelNex processes product information to authenticate users, enforce access, capture and recover recordings, generate and execute workflows, support human review, learn workflow-specific rationale when enabled, and maintain operational and security records.

Authorized members of your organization can access information according to their roles. Configured model providers, integrations, infrastructure providers, and other processors may receive the information necessary for the features your organization uses. For example, generating or reviewing a workflow can send recording content to the configured model provider, and executing it can send inputs to a connected application. Your administrator can identify those providers and their terms.

Precept Labs may process information needed to provide contracted hosting, support, security, and legal compliance. A customer-managed deployment does not imply that all connected services run inside that customer’s network.

Data-use limits

We do not sell browser-recording data, use it for advertising, or use it to train general-purpose models. Processing by configured model services is for the requested ModelNex features, such as generating, reviewing, and executing workflows. Workflow-specific learning, when enabled, is distinct from training a general-purpose model.

Retention and deletion

Retention of submitted recordings, workflow data, operational records, and backups depends on the deployment’s configuration, customer agreement, and applicable retention requirements. There is no single retention period that applies to every ModelNex deployment. Contact your organization’s administrator for its retention and deletion process.

The extension keeps local recovery data so interruptions do not immediately lose a recording. Saved or discarded sessions and replay data are cleaned up by the applicable product lifecycle; browser or extension storage may also be cleared through browser controls. Removing local data does not remove server backups, submissions, or audit records. Recording access grants expire and the surface requests revocation on sign-out or unmount.

Requests and changes

For access, correction, export, or deletion of organization-controlled data, contact your organization’s administrator. You can also contact support@preceptlabs.ai so we can direct your request to the appropriate party. Some information may need to be retained to satisfy legal, security, or contractual requirements.

ModelNex is intended for organizational and professional use. We will update this page when these practices change and revise the date above.