The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.
About This OTS Dataset
This Off-The-Shelf (OTS) dataset provides a comprehensive collection of audio recordings featuring conversations between Hindi-speaking customers and call center agents within the banking sector. It is meticulously curated to enhance speech recognition and conversational AI models tailored to the unique dynamics of Hindi-language interactions in financial services.
Metadata Availability: Insights into Participant Details
Each recording is accompanied by detailed metadata, including:
- Age
- Gender
- Country
- State
- Dialect
- Domain
- Topic
- Call Type
- Call Outcome
This rich metadata enables informed decision-making during model training and development.
Audio Recording Specifications
- Audio Duration: 400 hours
- Format: WAV (ensures high audio fidelity)
- Sample Rate: Adjustable to meet specific project needs
- Recording Environments: Captured in diverse real-world settings
- Equipment: Standard call center devices used for authentic conversations
These specifications ensure optimal performance across a range of AI and ML development applications in the banking domain.
Insights into Audio Data
The dataset includes 400 hours of high-quality recordings covering a diverse range of banking-related topics. Key highlights:
- Created with native Hindi speakers across India
- Captures realistic customer-agent interactions
- Includes multiple accents, dialects, and demographic profiles
Dataset Transcription Details
Each audio file includes manual verbatim transcriptions in JSON format with:
- Speaker-wise dialogues
- Time-coded segments
- Non-speech annotations
This enables rapid development of automatic speech recognition (ASR) and conversational AI models focused on banking call centers.
License
Exclusively curated by Macgence, this dataset is available for commercial use, empowering AI development in the Indian banking sector.
Updates and Customization
- Regularly Updated: New audio added from various real-world contexts
- Custom Options Available: Including sample rate adjustments and tailored transcriptions
👉 Submit your request to discuss your requirements, get pricing details, and purchase the dataset.
Keywords
keywords: Hindi speech dataset, banking speech dataset, Hindi customer support audio, financial conversation dataset, call center speech data, Hindi ASR dataset, Hindi transcription dataset, Indian banking dataset, Hindi conversational AI, Hindi call recordings, speech recognition banking, customer service dataset Hindi, banking domain audio corpus, Hindi voice dataset, speech data for banks, Hindi customer interaction dataset, Hindi voicebot training, commercial ASR dataset, speech dataset India, Hindi audio transcripts
- Downloads last month
- 78