The Rocketlog Slack integration adds an on-call AI SRE into your Slack workspace. You can ask it to fetch telemetry evidence (logs, traces, metrics) around a deployment or a time window so you can triage incidents and root cause without leaving Slack.

​
What the AI SRE does

  • Fetch evidence by time — Ask for “what happened between 2:00 and 2:15?” or “show me errors in the last 30 minutes.” The bot retrieves logs, traces, and metrics from Rocketlog for that window and summarizes or links to evidence.
  • Fetch evidence by deployment — Ask for “what changed in deployment 1.2.3?” or “telemetry for the last deploy of checkout-service.” The bot uses deployment and instance context (see Log enrichment) to pull relevant telemetry.
  • Summarize and suggest — The AI can summarize failures, point to likely root causes, and suggest next steps, using the same correlation and analysis you get in the Rocketlog dashboard.
This helps on-call engineers get context quickly during pages and reduces context-switching to the Rocketlog UI when Slack is already the incident channel.

​
Setup

  1. Install the Rocketlog Slack app — From the Rocketlog dashboard (or your admin’s link), add the Rocketlog app to your Slack workspace. Authorize the requested scopes (e.g. channels, chat, commands).
  2. Connect your workspace — Link your Rocketlog account/org to the Slack workspace so the bot can access your telemetry.
  3. Invite the bot — Invite the Rocketlog bot into the channels where you run incidents (e.g. #incidents, #on-call). You can also DM the bot for private triage.
  4. Configure permissions — Ensure the bot has access to the projects/environments you want to query (e.g. production). Your Rocketlog admin can restrict which data the Slack bot can see.
Exact steps may depend on your Rocketlog plan and how your org deploys the Slack app; follow the in-app setup flow or ask your admin.

​
Using the bot

  • In a channel — Mention the bot (e.g. @Rocketlog what failed in the last 10 minutes?) or use a slash command if configured (e.g. /rocketlog last 10m).
  • By deployment — Ask for evidence around a specific deployment or release; the bot uses deployment tags and time ranges to fetch the right data.
  • By service — Ask for a specific service (e.g. “checkout-service errors in the last hour”) to narrow results.
The bot responds with a summary and, where applicable, links to the Rocketlog dashboard for the same time window or deployment so you can drill in.

​
With alerts

When you use Rocketlog alerts, you can send notifications to Slack. The on-call AI SRE can then be asked to “explain this alert” or “get evidence for this incident,” using the alert’s time range and affected services to fetch and summarize telemetry. Together, smart alert grouping and the Slack bot help you go from “alert fired” to “root cause” without leaving Slack.