Training voice-to-text software to work in every day settings is a full time job.

Accents

Tali began its journey with hundreds of audio files from different accents reading medical notes, and we used them to fine-tune our algorithm. We continue to gather audio files from various sources to strengthen the quality of the dictated notes. We are committed to continuously improving our product's understanding of accents with contributions from the medical community.

We welcome your feedback on how well Tali interpreted your recent dictation.

Language

Tali is working on language translations in two areas:

Medical Language

Tali's language model is trained using state-of-the-art Natural Language Processing (NLP) algorithms on a multitude of medical resources and guidelines. As a result, it picks up medical terminology and abbreviations a lot more intelligently than a non-specialized voice-to-text solution.

Spoken Language

Tali started its training in English, but we are now moving into multi-language support. Please contact Tali to find out when your native language is scheduled to release.