How to process panel data - time series data consisting of a number of individuals with different number of time points each?

I want to fit a RNN of some sort to some panel data. All the examples I have seen assume just a single time series or several batches each with the same time dimension size.
In my case, I have data with a time dimension, and an individual dimension. Each individual forms a time series. Each individual's time series is of a different length of time.