Counterfactual prediction for non-linear relationship

[davidmosca]

Hello,
I am trying the solve the following problem. Let us assume this simple causal graph:
T -> R -> O
T -> V -> O
C -> V
C -> R
C -> T,
where C is a confounder, T the treatment and O the outcome, with O=R*V (this last relation is deterministic). All variables are continuous.
The relationship between O and counterfactual T is of shape unknown, non-linear, most likely smooth.
What I would need is the value of (counterfactual) O for a range of counterfactual T's, for a fixed confounder C=c (I am not interested in ATE or CATE).
Is it possible to solve this problem with EconML (and/or DoWhy)?
Thanks.