Tota is a Mac dictation app that transcribes entirely on your Mac, so it suits any profession where what you dictate must not reach a third party, or where a glossary of your own vocabulary makes the difference between usable text and a mess.
Every page below answers the same three questions for one kind of work: what leaves the Mac and what does not, which workflows the app fits, and what to put in a glossary so names, acronyms and terms of art come out spelled correctly. Each also says plainly what Tota does not do.
What these professions have in common
Cloud dictation apps send your audio to a vendor's servers, often on to a third-party AI provider, and keep it under their own retention policy. For most people that is a fair trade. For anyone bound by privilege, patient or client confidentiality, source protection, an NDA or an ethics approval, it is a disclosure that needs justifying. Tota removes the disclosure: speech recognition, formatting and speaker identification all run on the Mac, there is no account, and the app works with the internet switched off. The security page lists every network call it can make.
The second thing they share is vocabulary. General speech recognition has never heard your client's codename, the drug you prescribe most or the acronym your team uses hourly. A glossary is a short list of those terms that biases the model while it transcribes, and Tota can switch glossaries automatically depending on the app you are dictating into.
