Monitors
Continuously evaluate and improve Fin's conversation quality at scale.
Monitors give you a consistent, scalable way to manage quality across your support conversations. Instead of relying on manual reviews or disconnected QA processes, they let you automatically identify, evaluate, and improve the interactions that matter most.
Prerequisites
To get the most value from Monitors, make sure the following are in place.
- Fin deployed: Ensure Fin is live and handling customer conversations so Monitors can evaluate real interactions.
- Defined quality standards: Establish clear expectations for tone, accuracy, compliance, and escalation handling to reflect in your scorecards.
- Pro add-on: Enable the Pro add-on to access Monitors and additional Insights features.
Video transcript
We know quality matters to you, but in reality, QA is often difficult to manage, inconsistent in practice, and high risk issues often surface too late. And on top of that, AI and human support are reviewed in entirely different ways. This makes it harder to maintain standards and even harder to improve things systematically. Monitors change all that. They give you full visibility into the conversations that matter most and a consistent way to score them right where your support happens. You decide what you want to monitor, every Fin conversation, escalation issues, or legal and security conversations you know you can't afford to miss. When you create a monitor, you set the criteria to automatically flag conversations, and you can use both natural language descriptions and structured filters to do this. Then you attach scorecards to define how they're evaluated. Think of it this way, your monitor chooses what gets reviewed and its scorecard defines what good looks like. Your scorecard reflects your standards, things like escalation handling, accuracy, tone of voice, compliance, and more. Now you define what passes, what fails, and where improvements are needed. And you choose which criteria can be scored by AI or by a teammate when human judgment is needed. And once your monitor is live, you can manage everything in one place. The conversations that matched their score and who's reviewing them. And you can set up alerts to see matching conversations in real time and get detailed reports to track how QA is driving both Fin and team performance. So forget ad hoc sampling and forget spreadsheet based QA. Monitors give you a repeatable system to ensure quality that scales with volume.
More on Custom Scorecards
Custom Scorecards define how each conversation in a Monitor is evaluated, giving you a clear and consistent standard to meet. You can include criteria like accuracy, tone, compliance, or escalation handling, and decide whether each should be reviewed by AI, a human, or a combination of both. Once attached to a Monitor, the scorecard is automatically applied to every matching conversation, ensuring consistent evaluation without extra manual effort.
Scorecards also give you flexibility in how you manage reviews at scale. If AI scoring meets your standards, you can enable auto-review to reduce the need for manual checks, freeing your team to focus only on edge cases or failed conversations. This balance of automation and human judgment helps maintain high-quality support while keeping QA efficient as your volume grows.
