Poission with random effects

I have two levels in my poission regression model with random effects (children and family level) which predicts count data
after i get the exponents coefficient
can i say that
y(i) = exp(a + bx1(i) + bx2(i) + bx3(i))
or is it different when using random effects
thanks in advance