Robin handles first-tier support across email, chat, Slack, Teams, the web widget, and your Service Portal. It answers end users, works through problems with them, and resolves what it can without involving your technicians.
Robin handles first-tier support across email, chat, Slack, Teams, the web widget, and your Customer Portal. It answers end users, works through problems with them, and resolves what it can without involving your technicians.
Note: Robin was previously called IT Autopilot, and before that the AI assistant. See IT Autopilot is now Robin.
What is Robin?
Robin is an AI technician that interacts with your end users across multiple channels. It answers questions, guides users through fixes, and executes approved actions on their devices, so common IT issues are resolved before they reach your team. It is available 24 hours a day, works from the context you give it, and is available as an add-on.
What Robin can do
Answer and diagnose
Robin responds immediately, from the Service Portal, email, Slack, Teams, or a widget on your own site. It answers general IT questions and diagnoses device issues using the context it has about the user and their machine.
Robin responds immediately, from the Customer Portal, email, Slack, Teams, or a widget on your own site. It answers general IT questions and diagnoses device issues using the context it has about the user and their machine.
Act on devices and in the cloud
With access to your RMM and helpdesk, Robin can read the state of a device and change it: clear a queue, restart a service, free up disk space. It can also act in connected cloud services, for example resetting a password or updating a group membership.
Install approved software
End users can ask Robin to install or update software you have pre-approved, and it installs and configures it on request. Nothing outside your approved list is available to them.
Escalate with context
When Robin cannot resolve something, it escalates to a technician and passes on the full conversation, the diagnostics it ran, and what it found. Tickets can be assigned and reassigned between Robin and your technicians at any point, in either direction, without losing that history.
Notice problems early
Robin acts on signals rather than waiting to be asked. For example, when a user is locked out of their computer after too many failed login attempts, Robin notifies them by email or SMS and starts them on the process of regaining access.
[UNVERIFIED: the previous version specified three consecutive failed attempts as the threshold. Confirm the number or leave it general.]
Report on itself
Every Robin conversation and action is recorded, and AI Center > Analyze shows what Robin resolved, what it escalated, and which actions it used. See AI Center: Analyze.
Where end users meet Robin
In the Service Portal, by email, in Slack or Microsoft Teams, and through the web widget on your own site or intranet.
In the Customer Portal, by email, in Slack or Microsoft Teams, and through the web widget on your own site or intranet.
You choose which channels to enable and what Robin is allowed to do in them. See Robin: AI assistant for end users and AI Center: Settings.
What Robin changes
For end users, support is available whenever they need it, in the channel they already use, without waiting for a technician. Many requests are resolved in the conversation itself.
For your team, repetitive tickets stop reaching the queue. And when something does escalate, it arrives with the diagnostics already run and the cause often already identified, so a technician starts from a summary rather than from scratch.
How Robin works
Robin is not a chatbot answering from a script. It works through a problem the way a technician would: running checks, interpreting the results, and deciding what to do next.
To do that it draws on:
- User account data
- Diagnostics to read device state
- Actions to change device state and run remote commands
- Software installation
- Your knowledge base
- The Definitions you have written
- Playbooks, for workflows you want followed exactly
Internally it uses large language models wrapped in orchestration logic to analyse the request, assess the device context, and choose the next step.
When Robin cannot resolve something outright, it works down through what it does have:
- Step-by-step guidance from your end-user-facing knowledge base, where an article covers it.
- Your Definitions, for company-specific terms, software, and context.
- General IT knowledge from its underlying models.
If none of that resolves it, Robin escalates with the context your team needs to pick it up.
How Robin improves over time
- Your documentation and configuration. The more knowledge base articles, Definitions, Playbooks, and approved scripts you give Robin, the more it resolves. This is the change with the largest effect and the one you control.
- Technician feedback. Your team can give direct input on Robin's answers.
- Updates from Atera. Capabilities and performance improve continuously without any action on your side.
Note: Your data is not used to train any AI models, including Atera's. See Robin Security & Compliance FAQ.
