Verbamatic

Verbamatic transforms real-time speech to text with near-zero latency.

It’s an end-to-end, natively multi-lingual solution with a set of 42 languages and 60+ language models for creating meeting reports, starting from the acquisition of audio-sources, transformation to text, text real-time editing and report publishing.

The output of Verbamatic can be accessed through time-stamp and playing time.

Verbamatic can be easily integrated with your systems via:

• Internal or external audio interfaces

• XLR/USB mixers

• XLR dynamic/condenser microphones

• USB dynamic/condenser microphones

• DANTE (Digital Audio Network Through Ethernet) virtual audio interfaces

Verbamatic can be used in two different modes:

• Supervised Mode: is the human-in-the-loop approach where a human operator controls the regular functioning and can make real-time editing to guarantee the complete correspondence of the transcription.

• Unsupervised Mode: is the 100% automatic approach where the software automatically produces documents based on an accurate and fine-tuned language that guarantees automatic-yet-precise transcripts.

The solution is powered by Spoken Language Identification, Speaker Diarization and Speaker Identification features based on which it is possible to:

• Identify the language spoken

• Identify the speakers

• Classify the speakers