Been working on this project for the past 4’ish years as a side-project and while it’s not complete by any means, home-assistant integration for the most part is.
Her current name is Jenny and while this name is kinda special to me, I do realize it is a bad choice for a personal assistant.
Why you might care?
Solutions that existed when I started working on Jenny were nowhere near the quality/usability I needed them to be and/or relied on cloud infrastructure hosted by their respective creators.
While Jenny is not fully offline (most notable speech-to-text), she’s extremely modular meaning that any part of the core can be extended with an fully offline counter-part (f.e. googles speech to text engine with mozilla deepspeech. This has not yet happend because of accuracy reasons.)
Then again, everything that can be offline without hindering quality/performance is.
The aim of this project was to create a personal assistant that works as good as f.e. a google home or amazon echo while tightly integration with home-assistant and other projects (kodi was very important to me too).
Name suggestion requirements
Should not be an actual name (or at least one very very rarely used)
Should have a female (or at least gender neutral) ring to it
Should roll of the tongue easily (no “r”, “p” etc.)
High frequency letters like “i” (think “Siri”) might improve wake-word detection
Should not already be used by someone else for there AI project or be a famous name (“Mika” was a name I really liked but that’s apparently a nickname for a singer)
Final words
I appreciate any input and while I understand that most people would want to “see” her in action before getting “invested”, I want to wait for a first show-case until I have a final name and the source up on github.
Don’t put a female name as others. Be different. That female name could always be some people’s hated ex girlfriend name or worse could be the name of the wife. I don’t want my personal assistant to respond when i call my wife
You can never please everyone and there always seems to be a request to be able to change the wake word.
If you can implement that, then the project becomes independent from the wake word and you can call it whatever you want… (Within reason)
The wake word can be changed and the user will have to record himself saying it quite a few times for it to work properly.
Having an nice name from the get-go would be great though.
I’m slowly giving up though (This might be the flu speaking though)
“Dobby” (as in the house elf from Harry Potter) is what my wife and daughter came up with as their chosen wake word for ours. Would make a good name for an actual project.
This might help you… I was thinking ‘James’ or ‘Jeeves’ (some may remember AskJeeves) but there must be better names in the same butler/servant space. And so I ended up on this website… Thank you Google.
I’d also argue that it must be a name that requires no pronunciation instructions in any language (so, phonetically written and avoiding letters that are pronounced differently by different cultures) and have as few syllables as possible (“Hey, Google” is ridiculous in this sense).
The problem is that such names are also commonly chosen for kids I love the name “Nina” and am not aware of any different way to pronounce this by anyone, but it is indeed pretty common…as a Futurama fan, Leela/Lila would also be cool, and definitely far fewer people with a name like that.
As it stands there hasn’t been one where I had this “that’s perfect”-feeling though.
I feel like this isn’t something I should rush as changing it after open-sourcing would be a huge pain.
Until I have something I’ll just keep on working on the project, fixing bugs etc.
Impatient to see your project show up on github 
(This might be the flu speaking though)
I love the name “Nina” and am not aware of any different way to pronounce this by anyone, but it is indeed pretty common…as a Futurama fan, Leela/Lila would also be cool, and definitely far fewer people with a name like that.