ZembraTech
TunisiePublication : 18 janv. 2026
AI Voice Transcription & Emotional Intelligence System
- Stage
- 4-6 mois
- Présentiel
- Échéance : 1 févr. 2026
- Speech Recognition
- NLP & Similarité
- AI/ML Readiness
- Intelligence Artificielle / Deep Learning
- Biomedical Signal Processing
- Backend engineering
Description du poste
Core Features:
- Multilingual speech-to-text across diverse accents and dialects
- Emotion and sentiment detection from vocal patterns and prosody
- Real-time and batch transcription pipelines
- High-accuracy inference via model optimization and fine-tuning
- Continuous learning with feedback/evaluation loops
AI Integration:
- Develop an AI-driven voice transcription and emotional analysis engine that processes audio input and outputs accurate, language-aware transcripts enriched with emotional context. Mobile integration (e.g., React Native) is secondary and a thin client.
Tech Stack:
- AI/ML: TensorFlow, PyTorch
- Speech/NLP: Speech-to-Text, NLP Pipelines, Signal Processing
- Backend/Data: Python, Model APIs, Data Pipelines
Profile:
- Engineer or Bachelor degree in Computer Science