README.md

metadata

title: Call Center Analysis
emoji: 🎧
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860

This is a GPU-accelerated microservice for analyzing call center audio. It performs:

• Speaker Diarization (Pyannote 3.1)
• Speech-to-Text (Whisper Large-v3)
• Emotion Recognition (Wav2Vec2)
• Biometric Profiling (Age/Gender Voting System)
• Agent KPI Scoring & Customer Journey Mapping

Prerequisites

• Server: AWS EC2 g4dn.xlarge (or any machine with NVIDIA GPU + 16GB VRAM)
• OS: Ubuntu 22.04 (Recommended) with NVIDIA Drivers installed
• Docker: Docker Engine + NVIDIA Container Toolkit