You can do that of course, but why?
The world is flooded with small single board computers capableof running either android or Linux.
What is your PCB going to have what the others do not?
To get such a thing going, you’re looking at a 6 layer, maybe 8 layer PCB which is not going to be cheap in small quantities.
It’s got fine pitch BGA’s which are not very easy to solder, and outsourcing will be expensive for small quantities. DIY is difficult to get reliable.
So you’ve got a datasheet now. If it’s got “confidential” stamped in ugly red letters all over it, do you think you can get the processor itself (for some reasonable price)?
It is nearly impossible to compete with mass production.
And when you’ve got the hardware working, you have to fine-tune some Linux / android distribution, and do perpetual maintenance for it. Even some of the companies that make such (Linux) boards are struggling to have a solidly working linux distribution for it, or depend on their community to do so. And such a community needs a bunch of people to contribute to get to something stable, which implies a decent amount of PCB’s out in the wild (maybe several thousands?)
I just had a look at the Pine64 website and a bunch of their products are “out of stock” because they can’t get parts of the covid thing and global chip shortages.
I do not want to completely discourage you. Maybe you’ve got a good idea and is it doable too. I do want to encourage you to think about such factors before you put a lot of time and energy into a project like this.
For a starter, you can have a look at Olimex. They mostly use Allwinner chips, but they’re friendly to open source and have made their newer Linux computers in KiCad and the complete project on Github. You can find them via: https://www.kicad.org/made-with-kicad/