The details behind your account.
We store your email address, permanent user ID, display name, account status, and registration time to identify your account and provide sign-in and profile management.
Sign-in credentials
Passwords are stored as salted, one-way hashes, not readable plaintext. Email verification codes are stored as hashes with expiration times, attempt counts, and usage status.
Invitations & account records
We retain invitation codes, their availability, the original inviter, account relationships, and changes to those relationships to manage invitation-based registration and account administration.
When you use an optional account feature, the details you provide and related action history are used to operate that feature. Access to these records follows the permissions for that feature.
Information that keeps compute connected.
Creating an account alone does not inspect your computer. After you link a client, the service records device identifiers, hardware and operating system details, client version, connection heartbeats, and task status to manage device ownership and availability.
The server stores a hash of the device token. The client keeps its binding in a local identity file. Device and task records are associated with your permanent user ID.
Device name, operating system, processor architecture, CPU count and model, GPU names and indices, client version, connection times, and available task reports.
Request, account, device, worker and model identifiers; start and finish times, status, token counts, and error codes. Request fingerprints help identify repeated requests.
AI request content passes through the gateway to the configured model service for processing, and responses are returned to the caller. The platform's AI usage records contain metadata, not conversation text. Content handling by the model service depends on its configuration.
Keeping access under control.
Your browser uses a secure session cookie to keep you signed in. The server stores a hash of the session token, the associated account, request verification information, expiration time, and last access time. Local browser storage helps notify other tabs when an account session changes.
Network requests provide your IP address to the server. We use hashes of information such as IP addresses and emails for rate limits. Website access logs may include IP addresses, request times, paths, and connection information to investigate faults and prevent abuse.
Audit records cover administrative permission changes, selected sensitive-data access, device authorization and task-control actions, and support connection activity.
Who handles the information.
Platform administration
Authorized administrators can review account, device, invitation, and service records according to their permissions. An authorized agent can see current direct members' display names, account identifiers, masked emails, device summaries, and service records permitted by the current membership. The invitation page does not expose invited users' email addresses.
Email delivery
Verification codes and essential service emails are delivered through Resend. The recipient email address and message content are provided to that service. Messages include the code, its purpose, and expiration time, but never your sign-in password. Sending records help control frequency and volume.
Model processing
The configured inference service receives the content needed to process an AI request. The platform also uses model and worker information to route requests and associate usage with the connected device.
Understand remote access.
Remote SSH support is enabled by default after a client connects. Platform operators can execute support commands and access files permitted by the user account running the client. Support commands may therefore read local information.
The support tunnel uses a temporary device key. The server stores the authorized public key, port, expiration time, and support session audit records for verification and troubleshooting.
Disable support in the client and save your choice, or revoke the device authorization on the website. Exiting the management interface or stopping an AI task may leave an independent support connection active. Ending support does not automatically reverse system changes already made.
Run the client as an ordinary user where possible. The client does not automatically elevate privileges. Administrator support requires a separate root authorization.
Manage what you connect.
- Review your account and device information in the account center, and update your display name.
- Choose when to start or stop available AI tasks. Manage remote support separately in the client or through device authorization.
- Sign out to end the current browser session. Resetting your password invalidates previous sessions.
Disconnecting a device or ending a session does not automatically remove existing account, activity, or audit records. This page does not provide self-service account deletion.
noreply@slgang.com sends notifications only and is not a support or inquiry mailbox. Do not send passwords or verification codes to that address.
For device setup and access controls, see the account & device guide.