Self-tuning FoundationDB

Hello, we have started a project aimed to improve the performance of FoundationDB through optimization of its configuration parameters.
As a first step, we have applied state-of-the-art machine learning approaches employed to self-tune the configuration parameters of databases.
We have published online a paper that describes our work and our preliminary results, which is available here.

We would love to have your feedback, as well as suggestions on things to improve and next steps.
Thank you!