This seems like false reasoning. She is using an intentionally biased sample (the candidate pipeline in a narrow subspecialty of software engineering) to draw conclusions about the larger software engineering candidate pipeline, which actually has a higher representation of women - not high overall - but higher than both her subspecialty and more importantly, higher than the current representation of women at tech companies.
> If we increase the inflow of women into tech education, we will automatically increase diversity in hiring.
This is not enough. She would have us believe that companies are automata that will just adapt to the changing composition of the pipeline with no other action needed. It's an absolution of corporate responsibility for diversity (that perhaps reflects her self interest as a founder/executive). But companies are not automata, they are entities with their own inertial biases, and in the case of large companies, these biases often are often rooted in the communities from which the companies sprang.
Companies, especially major tech companies, are major cultural influencers - their businesses are based in part on the power of their cultural influence, and with that comes disproportionate responsibility to act, especially when society at large has not been successful enough at solving the pipeline problem.