Workflows

Agent Platform

Answer field issues in Slack and Jira

Let the agent take the first pass on support questions, where they already arrive.

Mention the agent in a Slack thread or on a Jira ticket. It checks the robot's data and your connected tools, then replies with what happened, the evidence and a next step.

When to use it

  • Field questions reach engineers who should be building.
  • Support waits on engineering to pull logs.
  • The same issues come back, and nobody remembers the last answer.

How to do it

  1. Connect Slack and Jira under Integrations. Alloy switches Slack on with you.
  2. In Slack, invite the bot to a channel with /invite @Alloy, then mention @Alloy in a thread.
  3. In Jira, type @Alloy or /alloy in a comment. On Jira Service Management, use an internal note.
  4. For a first pass on every new post or ticket, Alloy sets up a support role with you for that channel or project.
  5. Select Open Alloy on a reply to continue in the app.

Example

Fictional example.

At 02:10, a night-shift operator posts in the support channel. The agent answers before anyone on the team is awake:

Ask: "DLV-31 sat outside lift B for two minutes twice this morning, then went back to base. Seen this before?"

Agent: Both times the lift gateway did not answer the robot's call, and the robot gave up after its 120-second limit. DLV-27 had the same failure at lift B three weeks ago, closed as the gateway dropping calls after a network change. Next step: check the gateway's last restart. [Both missions linked]

Tips

  • Connect Confluence so the agent follows your own triage guides.
  • Ask the agent to link related tickets, so repeats are easy to spot.
  • Several issues at once are fine. Each thread and each ticket gets its own investigation, at the same time.

On this page