I find compounds with ring structures a struggle to determine synthetic routes towards because there always seems to be the issue of stereoselectivity in certain steps. To generate the stereocenter with the 2-methyl propylene substituent, probably would need to use some enantioselective enolate-type chemistry. Perhaps use an enantioselective Lewis base (catalytic) e.g. proline or some other secondary amine catalyst to make the enamine in situ (enolate analogue), and react that with an appropriate substrate containing the 2-methyl propylene backbone?
Also, I was thinking that maybe some allyl cation chemistry would be awesome in this case - the 2-methyl allyl acetate would be an ideal substrate for a Pd catalyst with an appropriate chiral ligand for enantioselectivity.
Grignard reaction would be best left to last to keep our ketone functionality (and looks like it'll work fine, Schlosser's base isn't necessary) - quick conformational analysis suggests that the Grignard should selectively form the product with the methyl and 2-methylpropenyl substituents cis.
I hope this helps somewhat!