As stated by Tim Porter in [his answer](https://mathoverflow.net/a/474375/113756), it all depends on what's the meaning you assign to the word 'concrete'.

Said that, possibly one of the most concrete (up to the limit of being applied and applicable) works of Alexander Grothendieck is, in my personal opinion, his work on (the algebrization of) Fredholm theory. His approach, described in the work [1] (and others references cited therein) seems inspired to an approach to abstraction I wrote about also in other posts on this and on the Math.SE sister site: according to this approach, "abstract" means "applicable in the widest possible context", thus in turn deeply concrete.

**Reference**

[1] Alexander Grothendieck, "[La théorie de Fredholm](https://doi.org/10.24033/bsmf.1476)", (French) 
Bull. Soc. Math. Fr. 84, 319-384 (1956), [MR88665](https://mathscinet.ams.org/mathscinet-getitem?mr=88665), [Zbl 0073.10101](https://zbmath.org/0073.10101).