The more labeled data you have, the more gains you can see in accuracy over time. As part of our on-premise implementations, we can provide Hotpepper, our end-to-end data labeling tool set.
Hotpepper allows you to build custom training datasets quickly and cost effectively while maintaining data quality and improves labeling efficiency by allowing users to submit audio to a hosted or on-premise Deepgram Speech Engine for automatic speech recognition (ASR) and transcription.
Administrators control the process, including availability and prioritization of files for labeling, as well as who may use the system.
If you are interested in adding Hotpepper to your on-premise implementation, please contact Sales.