Sliq maintains a minimum 92% word error rate quality benchmark. All our customers are periodically reviewed to ensure this level of accuracy is maintained. ...

Typical Accuracy Training Timeline*  *The following timeline is a representation of average accuracy performance and is subject to variance per customer...

Sliq’s ASR is used to provide closed captioning text for our video streaming service. Sliq’s ASR can also be used to provide verbatim text to create the fou...

Sliq’s ASR uses a combination of customer-provided text (corpora), audio files from historical meetings, and custom trained words identified by Sliq’s team ...

For new ASR projects, we recommend having 50-100 hours of meeting audio to establish a strong foundation for the ASR acoustic model. If this is not feasible...

Sliq’s proprietary solution is capable of using different backend systems to ensure access to the strongest performing results. Currently, Sliq’s solution u...

Sliq’s ASR typically involves a pilot project and set up a fee for new customers. A pilot is done to ensure the Sliq ASR is a good fit before committing to ...

Sliq’s ASR pilot is designed to cover the initial training period and set up of Sliq’s ASR service. There is a total of 4 training cycles allocated to achie...

Sliq ASR has a built-in silence detection feature. As such, running meetings with no audio should incur no additional cost. Sliq can provide a usage breakdo...

Sliq’s ASR can produce SRT files with timestamps for transcribed meetings. These same timestamps are used by our ASR transcription and training team to easi...