Discord Emotion Tracker — NLP Sentiment & Voice-of-Customer Analytics

Company: SupperTree, South Korea · Role: Software Engineer

An AI-powered application to analyze customer service communications on Discord, providing sentiment analysis and real-time language translation for actionable insights.

Key Contributions

• Captured and analyzed real-time Discord interactions, building a labeled dataset for sentiment analysis and emotion classification.
• Applied NLP via AWS Comprehend and HuggingFace Rust-BERT to detect emotional patterns across high-volume message streams — directly applicable to customer satisfaction and churn risk signals.
• Implemented real-time language translation (AWS Translate, English → Korean) enabling cross-market customer insight aggregation.
• Structured MongoDB storage for original messages, translations, and sentiment metrics — queryable for downstream analytics and reporting.

Companion Repositories

Related open-source Discord bot projects exploring similar architectures:

• discord-playdapp-bot — Rust · Serenity · MongoDB · Docker A high-performance Discord bot in Rust using Serenity for gaming community management. Handles tournament ticket exchange, player engagement tracking, and leaderboard functionality via slash commands.

• outfit-square — Python · Discord.py · MongoDB · Poetry Python-based Discord bot for user point management in metaverse gaming channels. Integrates MongoDB for persistent storage with multi-stage deployment and OAuth2 authentication.

Tech Stack

Rust · Serenity · Python · Golang · AWS Comprehend · AWS Translate · HuggingFace Rust-BERT · MongoDB · Oracle