About This App

This demo shows Q&A examples grouped by agent stage, but the production app is a batch pipeline — input a brand or keyword, five agents run in sequence, output is a packaged voice-of-customer report with sentiment breakdowns and action priorities.

The real 5-agent pipeline:
1. Listening Agent — scans social channels, collects mentions
2. Sentiment Agent — classifies positive / negative / neutral
3. Visualization Agent — builds sentiment charts
4. Export Agent — structures report-ready data
5. Packaging Agent — final report + recommendations

Dataset powering this demo: 500 coffee shop reviews (Jan–Mar 2024) across Twitter, Instagram, Facebook, and Google.

Email: yven81@gmail.com

Phone: +6014-9207099

Public Sentiment Collection Agent

Voice-of-customer pipeline for Malaysian F&B operators.
Sample case: coffee shop reviews — Q1 2024, 500 reviews, 4 channels, 5 product categories.

Pipeline Input: 500 reviews • 4 social channels • 5 categories • 334 unique customers • 5 agents • 15 sample analyses

Listening Agent Scans social channels, collects mentions and reviews

Sentiment Agent Classifies each review as positive, negative, or neutral

Visualization Agent Builds sentiment charts and ranked breakdowns

Export Agent Structures clean tabular data for the monthly report

Packaging Agent Writes the voice-of-customer report and action priorities

Click any question above to see that agent stage's output.