Project Overview
This SaaS tool tracks brand mentions across social media networks, customer reviews, and forums. It evaluates customer moods using natural language processing (NLP) pipelines, displaying live feedback streams with color-coded sentiment tags.
Key Features & Scope
Live sentiment feed updating via WebSockets (Socket.IO)
Keyword extraction highlighting trending problems or features
Automated support ticket generation for highly negative feedback
Multi-language support for international customer bases
System Architecture
React client connected to a NestJS gateway. Sentiment classification runs on a microservice wrapping Hugging Face Transformers models.
Client InterfaceNext.js / React
──────▶
Backend CoreNestJS API
──────▶
Database NodePostgreSQL & Redis
Database: PostgreSQL & Redis
Deployment: DigitalOcean Kubernetes
Performance & Vitals
96
95
98
100
Verified Production Metrics
Technologies Used
ReactNestJSHugging Face TransformersPostgreSQLSocket.IOTypeScript