Your environment and context
KeplerCrew supports deployment on a customer machine or private infrastructure, as configured. Tasks, conversations, project context, tool results, and run evidence are processed within the boundaries of that configuration.
Private deployment does not by itself mean that every operation is offline. External model endpoints and integrations may receive the context needed to perform an authorised request. Confirm those connections with your administrator.
Access and execution controls
The hosted application uses authenticated access. Customer-managed deployments have their own network and access requirements. Configure repository permissions, execution isolation, model connectivity, and administrative access for the environment in which the product runs.
Transport protection, storage encryption, credential management, backups, and retention must be evaluated for that deployment. This overview does not assert one encryption configuration or operational policy for every installation.
Reviewable work
Auto-pilot and Co-pilot are two modes of the same workflow. Co-pilot can pause at configured review points. Projects and runs provide context, activity, and recorded evidence that help a reviewer understand what happened.
Review gates and visible evidence support oversight; they are not a guarantee that every output is correct or safe to deploy. Inspect the changes and verification results before authorising consequential actions.
Connected providers
Model providers, hosting, storage, and other connected services depend on the selected deployment and agreement. Request the applicable provider list, processing locations, and data-processing terms before sending sensitive information through those connections.
See the privacy policy for the configuration-dependent handling of workspace records and privacy requests.
Security evidence for your review
Ask for the security documentation and evidence applicable to your proposed deployment. This public overview does not claim a certification, completed independent audit, guaranteed retention deadline, or service-level agreement.
Report a security concern
Send a private report to support@aichargelabs.com. Include the affected feature, safe reproduction steps, and potential impact. Avoid exposing another person’s data, disrupting services, or including credentials in the initial message.
For deployment-specific security questions or data-processing agreements, use the same contact.