TLDR

In under five minutes you can create an always-on watchdog that detects feedback anomalies, writes a root-cause summary, and alerts your team in Slack or email.

What is Quality Monitor?

Quality Monitor is Enterpret's first AI Agent. It continuously listens to every feedback stream in real time, surfaces statistically significant spikes, diagnoses the cause, and pings the right team—often before customers file a second ticket.

Perfect for CX, Support, Product, and Engineering teams that want to fix issues while they still affect a handful of users, not hundreds.

How it works behind the scenes

Listen 24/7: All feedback sources feed into an anomaly-detection layer.
Triangulate: Custom models score each spike for severity and relevance.
Explain: A root-cause-analysis model drafts a plain-English summary and delivers it on your preferred Slack channel or email.
Alert: Updates the alert as more customers are impacted.

Overview

Quality Monitor is an agent powered by Enterpret, designed to automatically detect, analyze, and alert teams about emerging quality issues in real time based on customer feedback anomalies.

How to setup a Quality Monitor

From the Agents tab on Enterpret, click on Quality Monitor to create a new Agent.

Here, you can specify two fields: a filter for the feedback that the agent should look at, and the destination where the agent should deliver alerts.

If no feedback filter is specified, the agent monitors all feedback by default for quality issues.

You can configure one or more emails and Slack channels to deliver alerts to. The agent delivers the alert as soon as it's identified, along with an RCA, feedback count, a trend chart, and other relevant details.

Note: Configuring at least one email or Slack channel as the alert destination is mandatory for creating a Quality Monitor agent.

Once you've configured the optional feedback filter and alert destination, click Create at the top right to save the agent.

You'll be prompted to give the agent a name. You can find this agent in the list of active agents later through this name.

That's it. You've configured and deployed a Quality Monitor agent to continuously keep you informed about quality issues reported by your customers' feedback.

Note: Members can edit agents they created without needing Admin access.

Alert reliability features

Duplicate-alert suppression: Deduplication reduces repeated notifications for the same underlying issue.
Alert grouping: Clique-based grouping helps keep distinct issues separate and prevents oversized alert groups.
Root-cause summaries: RCAs are derived from underlying user reports to improve diagnosis quality.
Clear alert titles and problem statements: Alerts describe the problem (instead of restating the solution) and use more specific titles.

Best-Practice Tips

Granular beats global: Create separate agents per product area to reduce noise.
Establish a triage ritual: Nominate an on-call owner for alerts.
Use the thumbs: Mark every alert "helpful" or "not helpful" to improve precision.