The Google engineer you described is valuable not because they know TensorFlow and ANN. Anybody who can learn computers can learn those things.
Then again you have no idea, you think they are interchangeable ... but they aren't....
The only way it could have anything to do with gender is if Google puts men into career paths that embed them in the uniqueness of Google, but put women engineers with the same skills in more outward facing positions, where they aren't really learning how the company works and grows internally.
Well first the women need to be engineers .... which as I pointed out the more equality you have the less STEM, WHY?
Also typically the great stride in engineering ie "value" aren't made by being social, or "knowing the business".
Having worked with Google now for 10years, I have only met one woman. But I am not a good metric, as I am on the highly technical side of voice and nlu work.
And if Google is doing that, they are discriminating on the basis of gender.
How would you calculate if Google is more of less bias? Its easy to do .. but you can't just use Google as your dataset.
Can any of you "science" people grok why the more gender equality you have the less STEM graduates?
If you can't, you don't know squat about what you are talking about, because you can't apply it across cultures.