Skip to main content

Sentiment Analysis with ShedBox Agent

Analyze customer sentiment across feedback, reviews, and support tickets.

Analyze Any Text Data

You: "Analyze the sentiment of these customer reviews"

ShedBox Agent:
✓ Analyzed 1,247 reviews

Sentiment Overview:
- Positive: 68% (848 reviews)
- Neutral: 22% (274 reviews)
- Negative: 10% (125 reviews)

Overall Sentiment Score: 7.2/10

Top Positive Themes:
1. "Easy to use" - mentioned 234 times
2. "Great support" - mentioned 189 times
3. "Time saver" - mentioned 156 times

Top Negative Themes:
1. "Slow loading" - mentioned 45 times
2. "Missing features" - mentioned 38 times
3. "Pricing" - mentioned 29 times

Use Cases

Customer Reviews

"Analyze sentiment in our App Store reviews"
"What are customers saying about the new feature?"
"Compare sentiment before and after the update"

Support Tickets

"Categorize support tickets by customer emotion"
"Identify frustrated customers who need attention"
"What issues are causing the most negative feedback?"

Social Media

"Analyze brand mentions on Twitter"
"Track sentiment around our product launch"
"Compare our sentiment to competitors"

Survey Responses

"Analyze open-ended NPS responses"
"Categorize feedback by topic and sentiment"
"What drives detractors vs promoters?"

Example Workflow

You: [Uploads support_tickets.csv]
You: "Analyze customer sentiment and identify issues"

ShedBox Agent:
✓ Loaded 3,456 support tickets
✓ Analyzing text content...

📊 Sentiment Analysis Results

Overall Distribution:
- 🟢 Positive/Satisfied: 42%
- 🟡 Neutral/Informational: 38%
- 🔴 Negative/Frustrated: 20%

Trend (Last 30 Days):
[Line chart showing sentiment over time]
⚠️ Negative sentiment increased 5% after Oct 15 update

Topic Analysis:

| Topic | Volume | Sentiment | Trend |
|-------|--------|-----------|-------|
| Login issues | 234 | 🔴 -0.7 | ↑ 23% |
| New UI | 189 | 🟡 0.1 | New |
| Billing | 156 | 🟡 0.2 | → |
| Speed | 142 | 🔴 -0.5 | ↑ 15% |
| Features | 98 | 🟢 0.4 | ↓ 5% |

Urgent Attention:
23 tickets marked as "angry" or "frustrated" requiring escalation

You: "Show me the tickets about login issues"

ShedBox Agent:
Login Issues Analysis (234 tickets):

Common Patterns:
1. "Password reset not working" - 67 tickets
2. "2FA issues" - 45 tickets
3. "Session timeout" - 38 tickets
4. "SSO not connecting" - 28 tickets

Sample Negative Ticket:
"I've been trying to log in for 2 hours! Password reset emails aren't
coming through. This is unacceptable for a paid product."
Sentiment: -0.9, Urgency: High

Recommendation: Investigate email delivery issues with password reset

Generated Pipeline

Sentiment analysis becomes automated:

data_sources:
tickets:
type: csv
path: support_tickets.csv

ai_interface:
provider: anthropic
prompts:
analyze_sentiment:
system: "You are a sentiment analysis expert."
user_template: |
Analyze the sentiment of this support ticket:

{{ticket_content}}

Return:
- sentiment_score: -1 to 1
- sentiment_label: positive/neutral/negative
- topics: list of topics mentioned
- urgency: low/medium/high
- key_phrases: important phrases

processing:
transform:
- operation: ai_enrich
source_field: ticket_content
prompt: analyze_sentiment

aggregate:
group_by: [topic, week]
metrics:
- avg_sentiment: avg(sentiment_score)
- ticket_count: count
- negative_ratio: count(sentiment_label == 'negative') / count

output:
type: file
path: sentiment_report.json

Analysis Features

FeatureDescription
Sentiment ScoreNumeric score from -1 (negative) to 1 (positive)
Topic ExtractionAutomatic identification of topics discussed
Entity RecognitionIdentify products, features, people mentioned
Urgency DetectionFlag tickets needing immediate attention
Trend AnalysisTrack sentiment changes over time
Comparative AnalysisCompare sentiment across segments

Get Started

Understand your customer sentiment at scale.

Try ShedBox Agent →