So if you switch on all three, you are at 25 cents per hour of recorded audio.

That is Assembly AI's current published pay-as-you-go pricing.

Your workload is Monday, Friday, 12 hours a day, 5 or 60 hours.

Saturday, Sunday, 2 hours a day, 2 or 4 hours.

So 64 hours a week.

At 25 hour, period.

Audio cost, normal weekday, 12H, $3 a day.

Saturday or Sunday, 2 or 5 a day.

Full week, 64H, 16s a week.

Average month, 2 roughly 77.3 urge, 69.33 a month.

Average year, 3,328.832 a year.

The monthly figure uses 52 weeks, 12, rather than pretending every month is exactly 4 weeks.

So yes, roughly $70 a month for your expected BRI workload,

including their best async transcription, diarization, and their speaker identification feature.

That strikes me as remarkably inexpensive for what you're doing.

But there's a crucial distinction about speaker identification.

This is where I need to sharpen my previous answer.

Assembly AI's speaker identification is not quite the persistent,

biometric voice recognition technology.

It's the communication system you've built in BRI.

Assembly AI can...