Caryvane Caryvane Data lifecycle management

One platform, end to end

Machines, clouds, mailboxes — your whole data estate, one platform.

Backup is where most people start, but it is one job among many here: application-consistent servers, archiving and retention, cloud-to-cloud migration, a mounted cloud drive, and Microsoft 365 and Google Workspace — every stage of your data's life, from one console.

Most jobs run on an agent installed on the machine. A cloud-to-cloud job can do the same, or run on a hosted worker instead — set per job, so you decide whether the data passes through your own hardware or never touches it. Open whichever of these you came for.

Where it can put the data Talk to us

File backup

Scheduled or continuous, to whichever storage you have pointed it at.

  • Full and incremental runs
  • Real-time watching for changed files
  • Version history and retention
  • Restore to the original path or elsewhere
  • Scripts before and after a run
  • On Windows, files that are open and locked are still captured

Servers & applications

Whole applications captured consistently, not just their files — the application is asked to settle before the copy is taken.

  • Hyper-V virtual machines, application-consistent while they run
  • Incremental VM backups on Server 2016 and later — only the blocks that changed travel, yet every backup is a complete disk image
  • Microsoft SQL Server databases, captured copy-only so your own log backups and point-in-time restores are never disturbed
  • Active Directory on a domain controller
  • Same storage, schedules, retention and version history as every other job

Cloud drive

Mount object storage as an ordinary drive and work from it directly.

  • Files appear in Explorer and Finder, and open in any application
  • Fetched on demand rather than synced wholesale
  • Local cache, so what you have opened stays quick
  • Files far larger than the disk they appear on
  • Writes queue and upload in the background

AI integration

Ask an assistant about your backups, or have it run one — from Claude, ChatGPT or VS Code.

  • An MCP server for Claude and VS Code Copilot
  • A Custom GPT Action for ChatGPT
  • Personal keys, issued and revoked by each user
  • Four permission tiers, from read-only upward
  • Read the dashboard, jobs, agents and run history
  • Start or cancel a run; create jobs; mount a drive

Cloud to cloud

Back up one provider straight into another, and choose what the data passes through.

  • Any supported storage as the source, any as the destination
  • Azure Blob to Amazon S3, S3 to SFTP, SharePoint to a NAS
  • Run it on your own hardware, or on a hosted worker — per job
  • Whole containers or a single prefix within one
  • Same schedules, retention and version history as any other job

Microsoft 365

Mail, OneDrive and SharePoint copied out to storage you hold.

  • Incremental by change token, not a full sweep each time
  • Archive older mail out of the mailbox
  • Restore individual items or whole mailboxes
  • Runs from the same console as everything else

Google Workspace

Gmail, Drive, Calendar, Contacts and Shared Drives copied out to storage you hold.

  • Connected once for the whole domain, not per user
  • Incremental rather than a full sweep each time
  • Archive older mail and Drive files out to a set you keep
  • Restore mail and Drive files back to the user
  • Runs from the same console as everything else