Question: How do I numerically solve for a root, involving a procedure?

I'm asking this because I'm sure I'm just missing something.

If I solve a DE with numerics, i.e., dsolve(...,numeric), I get a procedure, say . I want to answer: if I have a point  that I know, what are the root(s) of ? In other words, solve for .

I thought this would just involve something like fsolve or solve/evalf, or perhaps a root finding package, but I can't figure it out. What I want to avoid is having to write a procedure to hone in on the root myself, since my code is already quite slow.

Any help would be greatly appreciated.

EDIT 1: Here is my MWE (or, what should work in my mind but doesn't). The ODE isn't important, it's just an easy example.

restart;

ode := diff(f(x),x) = f(x);

initial_condition := f(0)=1;

a := 1;

h := dsolve({ode,initial_condition},numeric);

solution := fsolve(h(x) - rhs(h(a)[2]) = 1);