It's been interesting to follow this game and compare it to ours. In our game if you play either 7.a3 or O-O and then c5 chances are roughly equal. My plan was then to castle long and attack the kingside, a version of the old-style main line of the Steinitz. Siro plays a sharper more modern line that I'm not familiar with, but I think your first misstep was 9. . . Bxc5. I think if you take with Nxc5 instead (hitting d3 if white still wants to develop the bishop there) and then castle, the game has equal chances.