---
name: "Sentiment Analysis"
description: "Plugin that analyzes customer message sentiment in real-time, surfacing emotional signals to help agents respond with appropriate tone and urgency."
type: plugin
roles: ["Customer Success", "Operations"]
activities: ["Analytics", "Support"]
tags: ["Analytics", "Sentiment", "NLP", "Customer Experience"]
owner: "Analytics Integration Team"
version: "1.8.0"
sourceUrl: "https://github.com/qiscus/sentiment-plugin"
compatibility: "Qiscus Platform >= 3.0"
---

# Sentiment Analysis

> Plugin that analyzes customer message sentiment in real-time, surfacing emotional signals to help agents respond with appropriate tone and urgency.

## Overview

Sentiment Analysis scores incoming customer messages for positive, neutral, negative, and frustrated sentiment in real-time. Agents see sentiment indicators alongside conversations, enabling proactive de-escalation before situations deteriorate.

## Capabilities

| Capability | Description |
|---|---|
| Real-time Scoring | Scores every incoming message for sentiment on a 5-point scale |
| Frustration Detection | Flags conversations showing escalating negative sentiment |
| Trend Visualization | Shows sentiment trajectory across a conversation timeline |
| Batch Analytics | Aggregates sentiment trends across teams, products, and time periods |
| Alert Rules | Triggers notifications when sentiment drops below configurable thresholds |

## Install Steps

### 1. Install via Qiscus CLI

```bash
qiscus plugin install sentiment-analysis
```

### 2. Configure alert thresholds

```bash
qiscus config set SENTIMENT_ALERT_THRESHOLD=-0.6
qiscus config set SENTIMENT_NOTIFY_CHANNEL=slack
```

### 3. Verify installation

```bash
qiscus plugin list | grep sentiment
```

Need help? Contact the Analytics Integration Team or check the Qiscus documentation.