Endpoint management
In productionMonitoring, scripting, patching and remote access from an agent Atlas builds and signs itself, plus a support tray the end user can chat to a technician through.
Measured: 307 managed devices, 2,014 jobs run and 3,346 alerts raised in production.
What it does
- Hardware and software inventory, health and alerting per device, mapped to the client that owns it.
- Scripts and remediations run through a job queue, with output captured against the device.
- Patch rings with an approval gate, plus third-party application patching through winget and brew.
- Remote access through AnyDesk, provisioned per device.
- A support tray on Windows, macOS and Linux: chat to a technician, raise a request, see who is helping.
- Software catalogue policy — required, approved and banned applications, enforced nightly.
- Unwanted software detection against a PUP and adware library, with removal routed through the approval flow.
What it does not do
- The macOS support tray is built and notarized but NOT PUBLISHED yet, so Mac users get the agent without the tray. Windows and Linux have both.
- Deleting a device in Atlas does not uninstall the agent from the machine. That is a deliberate separation, and it surprises people.
- Patch compliance reporting is per-ring and per-device. There is no regulator-shaped patch attestation report yet.
What it replaces
- A standalone RMM
- A separate remote-access subscription
Included in the single per-user price. No per-module or per-endpoint add-ons.
Read the guide
How this is actually set up and run, in the documentation.