What: Detect student emotions (frustrated, confused, excited, etc.) from text and voice, then adapt teaching style automatically.

Tech stack:

• FastAPI (REST API endpoints)
• LangGraph (workflow orchestration)
• HuggingFace Transformers (RoBERTa, BERT) for sentiment/emotion
• Wav2Vec2 for voice emotion recognition
• librosa for audio feature extraction
• Qdrant to store emotion history
• Streamlit/Plotly for emotion analytics dashboard

Key skills: NLP, audio processing, adaptive systems, ML model integration

📧 Pour postuler: contact@cloudiasys.com