Call Center Audio: 13,000+ Hours of US Customer Service Conversations

UniDataPro provides 13,000+ hours of real-world call center audio recordings featuring over 90% unique speakers. The collection includes time-stamped transcripts designed for training speech recognition and speaker diarization models in the customer service domain.

Use Cases

• Training Automatic Speech Recognition (ASR) models using audio and time-stamped transcripts
• Developing speaker diarization systems to distinguish between 90%+ unique speakers
• Analyzing customer interaction patterns within US-based call center audio segments

Strengths

• 13,000+ hours of audio data
• 90%+ unique speaker diversity
• Includes time-stamped transcripts for model alignment

Limitations

• Restricted by CC BY-NC-ND 4.0 license prohibiting commercial use and derivative works
• Geographic focus is limited to the US region

Provenance

Source

UniDataPro

Collection Method

Real-world call center recordings

Freshness

Last updated March 2026.

Geography

United States