SB here, founder of Onepin. Our launch goes live on Saturday at 12:01 AM PT. I want to use this thread to learn how you work, not to pitch.
Here is what we keep seeing. A team generates 1,200 lines of AI voice for a product video, an ad campaign or a course. The voice sounds great. Someone listens to the first twenty lines, nods, and ships the rest on trust. Line 1,200 reads the product name from its spelling, and the customer is the first one to hear it.
We spent two years at Podonos measuring how AI voices sound, with 150,000 human listeners doing the rating. The voices got the delivery right. What kept failing was names, prices and dates, and the only check most teams had was one person with headphones.
So, honestly:
1. Do you listen to all of it, some of it, or none of it before it goes out?
2. What has slipped through? A brand name, a number, a date, your own company's name?
3. If a check existed that caught those before you published, what would it have to get right for you to trust it?
I'll answer every reply here, before and after launch day. If you want to hear what the check catches, onepin.ai runs one line without an account.
The drug terms in the 4M-word dictionary caught my eye. I'm building a voice AI that phones elderly parents every day, and the failures that hurt trust most are exactly this kind: a medication name or a family member's name said slightly wrong, and an older listener stops believing the call is from someone who knows them. Question: does the pronunciation check work on short, live-generated lines (a few seconds, no time to re-render), or is it designed for batch scripts like videos and courses? And can a user pin a custom name, like a grandchild's, so it is always said right?
@igorgurovich Great question: for a full transparency, we have been focusing on the accuracy while sacrificing the latency. Our roadmap is to make AI accurate first and then to make them fast. I have been hearing from numerous voice AI agent companies that the errors on critical terms are deal breakers, we are working toward making them possible. Technically, it is called a distillation (or teacher-student network) where a big neural network teaches a small one that typically runs fast, we are on this. Please contact me directly.
Getting names, prices and dates pronounced correctly is the small detail that makes AI voiceovers usable in real work, since a garbled client name is an instant give away. Checking against a four million word pronunciation dictionary and scoring each line is a thorough way to catch the misses. Does it let you save the corrections so the same name is right next time?
@karimbenkeroum Oh great idea. We spent months for building 4M dictionary, but it doesn't cover all the new words. So our agent instantly looks for pronunciations from web, YouTube, etc, and uses them for correct generation, and most importantly remembers them. What is your use case?
the naming problem is real, we deal with the same thing on live calls at Dial. the difference for us is latency, we can't check and re-render before the line ships because there's no "before" at all, it's a live turn. is Onepin scoped to pre-rendered scripts where you have time to check first, or could the pronunciation dictionary and scoring run fast enough to sit in a live TTS path too
@galdayan great question. I understand your concern: on the voice AI agent call, you want to make it work. It is on our roadmap. We understand your problem, we decided to first solve the error issue first and then address the speed issue. Small hint: distillation. Our small network (under training) will learn the big network that is currently working as a part of the current Onepin service. Stay tuned!
Hi, great product. Planning to support Turkish language?
@aytekin_meral Yes. Turkish is one of the languages we will support next. What is your use case?
@soohyun Voice AI checking names and prices before speaking sounds really useful for avoiding mistakes
@shivam_kushwaha16 Thanks. Mainly two things: making the script to be well spoken by AI, which is called normalization. Also, checks for the pronunciations and auto fix, where our internal models reviews the pronunciations and fix them. Please try!
@soohyun I will definitely check it out properly and try it myself
@shivam_kushwaha16 Of course. For any suggestions or questions, directly contact me.
Passed this straight to our video editor we do a lot of multilingual dubbing and localized numbers always ness up the takes.
@alira_salu Great. If you have any issues, please directly contact me. Question: which languages do you need to cover?
I usually hate having to listen through every line of AI voiceovers for errors so the line scoring feature is a huge time saver for our team.
@evelyn_swaim Awesome. Hope Onepin helps you waste time. Please contact me directly for any suggestions.