This is great and including the data generating model with a regressor made it really clear, thank you!
The only follow up question I have - any recommended learning resources or books that would help me start coming up with my own solutions like this in the bayesian / time series / GLM space? I’ve read Rethinking, but it doesn’t have much on Time Series (and I’m a little weak with time series anyway tbh).
I would say the biggest area where I don’t understand the theory is how using the latent variable that is a random walk accounts for the autocorrelation