Support Agent

Answer field issues without pulling in your senior engineers. The agent works in your Slack support channels and on Jira tickets, investigates each request against the robot's data, and replies with what happened and what to do.

What it is

A support agent for your robot fleet. It answers field issues in Slack and Jira with the robot's own data, before they reach your engineers.

How Alloy does it

It answers where requests arrive. A new post in a support channel, an @Alloy mention, a direct message, or a new ticket in a Jira project you choose.
It investigates the robot, not only the ticket. It finds the run, reads the recording around the reported time, and checks your triage guides, earlier cases and the metrics you connect.
It replies with evidence. What happened, the moment in the recording, a likely cause, the next step, and what it could not confirm. Its replies on a Jira ticket post directly; other ticket changes wait for a person.
It keeps the record. Each request becomes a case with its outcome. Optional follow-ups ask whether the fix held when a thread goes quiet. Alloy sets up the support roles with you.

Ask in the Alloy app, in Slack, on a Jira ticket, or from your coding agent.

Example

Fictional example: the robots, the depot and the numbers are invented.

A depot supervisor calls your support line. Your L1 support specialist posts in #field-support:

"Cleaning robot CR-221 at the Riverbend depot stopped mid-route, operator says the brush won't restart. Robot is otherwise fine."

The agent replies in the thread:

What happened: at 10:41:07 the brush motor controller reported an over-current fault and the robot stopped the brush, then the drive, as designed. The brush motor current rose from about 9 A to 31 A over 1.5 seconds before the fault. Drive speed and floor slope were normal. [Open the moment in Replay]

Seen before: CR-221 had two similar over-current faults last Tuesday, both cleared after a restart. No other robot at Riverbend shows this fault in the last 30 days.

Likely cause: something wrapped around the brush spindle (the current rise is gradual, not a step). This matches "Brush over-current" in your Cleaner triage guide. The brush controller settings are the same in the last two releases.

Next step: ask the operator to power down, check the spindle for cable ties, tape or debris, clear it, and restart. If the fault returns within a day, escalate to hardware with this thread.

Not confirmed: I cannot see the spindle. The cause is inferred from the current profile and your triage guide.

The specialist sends the steps to the operator. The robot is cleaning again in 20 minutes. No autonomy engineer was paged. The next day the agent asks in the thread whether the fix held, and records the answer.

How to start

Ready to start. Talk to us. Alloy sets up your workspace with you, connects your data and tools, and the agent takes the first pass on live requests from the start.

Want proof first. Send a handful of tickets you have already closed, with their logs. Alloy answers them without seeing your team's replies, so you can compare.

Just looking. See the agent work on sample robots in Try Alloy.

What we need from you: access to your robot data, the Slack channel or Jira project, and your triage guides.

Outcomes for your team

Field issues get answered in hours, not days. An investigation starts the moment a request arrives, not when an engineer is free.

"Last year at this time, I was spending a day or more investigating issues and support tickets. … Now it's usually less than an hour."

Mitch Galea, Senior Autonomy Engineer. How SwarmFarm is scaling its robot fleet

Your engineers stop being the first responder. The agent takes the first pass, and your engineers join when the evidence is in.

"Our engineer would respond first and help out, … Now Alloy is the first to respond and help out."

Sophie Lane, HRI and AI Engineer. How Andromeda uses Alloy

Fewer tickets reach engineering. Support staff and operators solve more on their own, with the evidence in the thread.

"Tickets are stopping before they reach us, … People are solving their own problems."

Tom Brodie, Head of Autonomy. How SwarmFarm is scaling its robot fleet

Root Cause Analysis: what engineers do with an escalated case.
Fleet Monitoring: catch an issue before the field reports it.
Slack, Jira, Confluence