Incident agent on Slack: internal MVP to test capture and tickets.
Ignacio De La CubaAI Agents
Context & Problem
Opening and tracking rental incidents relied on informal channels or forms disconnected from how the team works day to day. That made it harder to capture context in conversation, route cleanly into the ticket pipeline, and validate data quality before exposing the flow to tenants and partners. Without a controlled MVP, we risk shipping on the final channel without having tested capture quality or thread-based context.
Solution
We shipped the first version of the incident agent on Slack: an internal MVP where the team opens and tracks incidents by chatting with the bot in the channel where they already work.
- Thread-aware conversation as incident context.
- Output wired to the ticket pipeline (routing and status in sync).
- Same agent core as the rest of the platform. Slack is our test surface; WhatsApp for the external ecosystem remains on the roadmap.
Expected impact
- Less detail lost from “something broke” to a ticket with useful fields.
- Real validation of the flow with the team before we expose it outside Slack.
- Lower risk when we open the same flow on the mobile channel with tenants and partners.