Quest Game Tuner's auto tuning is actually per-game as well!
You pick your preference of performance, visuals, or battery, and the rest is filled out automatically.
I'm a professional data analyst by day, so I decided to fully automate the process.(Technical stuff incoming!)
I collected a ton of my own data for profile apps and games on my own devices, then trained a Torch model which can evaluate a game's performance profile based on info from it's metadata and store page, viaoculusdb. The AI's outputs go through additional post-processing to refine it's outputs for your device and preferences.
The result was a tiny model that can generate dozens of high-quality tuning profiles per second, and can figure out the best settings for a new game before it even finishes opening.
Unlike qgo, it's totally on device and doesn't rely on user data.
Interesting, so a nice bonus is that you have immediate access to profiles on new game releases rather than waiting for the community to build/submit profiles for new games in QGO. Assuming you did side-by-side comparisons of popular games with your product and QGO, is the end result similar across games or does one provide a better experience than the other?