SLA-breach root-cause investigator agent

This workflow runs an autonomous post-mortem on every SLA miss. When Intercom marks a conversation as breached, an agent reconstructs what happened — the full reply timeline, how deep the queue was at the moment of breach, and which agents were available — reasons about whether it was a staffing, routing, or complexity failure, and publishes a written root-cause analysis. It then files a follow-up task so the gap gets closed, not just logged.

Use it when you want continuous improvement on SLA performance rather than just alerting, and your team needs a documented record of why each breach occurred. It is the learning loop behind the predictive routing workflows.

1. 1An Intercom SLA-breach event triggers the flow.
2. 2The agent retrieves the conversation timeline and metadata from Intercom.
3. 3It reconstructs queue depth and agent availability at the breach moment.
4. 4The agent reasons over the evidence to classify the root cause and recommend a fix.
5. 5It writes a structured root-cause summary to the Notion incident log.
6. 6It opens a Linear task assigning the recommended follow-up to the support lead.