Skip to main content

Integration · Slack

Slack Alerts The Moment Users Hit Errors

Real-time alerts when sessions break, with one-click replay opens. Triage faster than your monitoring stack can ping. No more "user said something on Twitter".

What you get

Real-time alerts

A high-frustration session, a Stripe API error, a checkout abandonment — alerts fire within seconds of the event in the user's browser.

One-click replay opens

Every alert message has a "Open replay" button. Click it; you're watching the session in your browser. No app-switching, no copy-paste session IDs.

Channel + DM routing

Route by team, by error type, or by customer tier. High-value enterprise customer hits a bug? It pings #cs-vips. Anonymous user? Lands in #qa-firehose.

Smart deduplication

Identical errors fire one alert with a counter, not 50 alerts. Stack-trace fingerprinting is automatic; custom dedup keys are supported on Pro+.

Setup in four steps

  1. 1

    Connect Slack workspace

    Slack OAuth — admin approves once. The Relyv app appears in the workspace; channels can opt in individually.

  2. 2

    Choose channels

    Add the Relyv app to the channels you want alerts in. Different channels can receive different alert types.

  3. 3

    Define alert rules

    Frustration score above X, specific console error, page path matches Y, customer tier is Z. Rules combine with AND / OR; preview matches against historical sessions.

  4. 4

    Replays land in your inbox

    Alerts appear with the AI summary, severity, replay link, and one-click ticket-creation button (Linear / Jira / GitHub).

Implementation details

Alert volume control

Two safety knobs: a per-rule rate limit (e.g. "max 1 alert per 5 minutes for this rule") and a global per-channel cap. Both are enforced server-side. Rules with a backed-off counter show "12 more matches in the last 10 minutes" so you don't lose visibility, just noise.

Custom rule examples

"User on Pro tier hit a Stripe error" → ping #payments-vips. "Frustration score > 0.85 on /checkout" → ping #growth. "Console error containing TypeError on /signup" → ping #frontend with @here. "Any session over 20 minutes long" → DM the founder.

Deduplication

Default fingerprint = stack trace + browser version + page path. Identical events within a 10-minute window collapse into a single alert with a counter. You can set a custom fingerprint per rule (e.g. "user_id + error_message" if you care more about per-user repeats than per-error).

Frequently asked

How granular can alert rules be?

Highly. Match on frustration score, console messages, network errors, page paths, page tags, custom user properties, AB-test variant, browser, OS, and combinations of all of the above. Rules are versioned; bad ones can be rolled back.

Can I receive alerts as a DM instead of in a channel?

Yes. Per-rule routing supports channels, DMs, or webhooks. Common pattern: critical alerts DM the on-call, info alerts go to a channel.

How is identity mapped between Slack and Relyv?

Slack handle → Relyv workspace member, mapped at OAuth time. If you want to mention specific users in alerts (@here is per-channel), the mapping is in workspace settings.

What about Slack Enterprise Grid?

Supported. Org-level admins approve; workspace admins handle channel-level config. Multi-workspace orgs can route per-workspace ("payments alerts to the #engineering workspace, marketing alerts to the #growth workspace").

Ready to wire up Slack?

Real-time alerts when sessions break, with one-click replay opens. Triage faster than your monitoring stack can ping. No more "user said something on Twitter".