I turned two launch comments into Conversation Profiles and same-language transcription
I recently launched TCVoiceAI, a Mac app that translates both sides of an online call from one computer. Two early comments asked for company-specific vocabulary and per-contact language preferences.
They sounded like separate feature requests, but they pointed to the same underlying problem: recurring conversations should not start from a blank setup every time.
I built Conversation Profiles around that idea. A profile can now represent a contact, team, or scenario and save its languages, conversation mode, context, translation preferences, and custom vocabulary. Each profile also keeps its own local History.
That work made another use case clearer. Sometimes people do not need voice translation at all; they need a searchable transcript of a call in the same language. Version 1.2 now supports same-language sessions in Original audio mode. Users can keep transcription only, which sends no translation request and uses no translated-word allowance, or optionally translate both speakers' transcripts into another language.
I also added live switching between Translated voice and Original audio during bilingual sessions. The difficult part was not the control itself, but changing audio-route ownership safely while the session was running. If the new route cannot be completed, the app restores the previous mode.
Calling-app setup was another source of friction, so the app now detects compatible installed apps and shows the relevant microphone and speaker steps. FaceTime can be configured automatically after the user grants Accessibility permission.
My main lesson was that the best feature requests often describe a repeated setup cost rather than the final feature. Solving that underlying cost produced profiles, vocabulary, separate History, and a cleaner path to transcription.
I am now trying to learn which workflow matters most in practice: two-way translated voice, same-language transcription, or reusable profiles for recurring calls.
TCVoiceAI: https://tcvoiceai.com/
I enjoyed the way you described how two seemingly different requests led back to the same underlying issue.
Those moments are often more valuable than the features themselves because they reveal how people are actually trying to use the product.
Thank you. That was the key insight for me too: users did not just want more settings; they wanted recurring conversations to remember them.
That makes sense.
The interesting signal will be whether users start expecting that continuity as part of the product experience, rather than seeing memory as just an added feature.
That expectation shift is usually where the real habit forms.
The stronger signal will be repeated profile use across recurring calls, not simply how many profiles are created. I’ll keep that distinction in mind as I gather feedback.
That distinction is exactly the interesting one.
Repeated use is what will tell you whether memory is becoming part of the workflow or just a nice feature.
I have a couple of thoughts on what could make that continuity valuable long-term, but I don't think I'd do it justice in a thread.
If you're open to it, what's the best email to reach you on?
Thanks, I’m open to hearing your thoughts. You can reach me at support@tcvoiceai.com. Please include a little context on what you have in mind.
Thanks! I’ve just sent it over.
Looking forward to hearing your thoughts whenever you have a chance.