Agentic forecasting

Technical deep dive + barebones implementation

I have been working with forecasting for a long time - I have seen a few hype cycles, watched fashions rise and fall… But one thing has remained unchanged: if you look closely, most problems in time series are solved with a function call:

y_hat = model(history)

It works, it is not that hard conceptually, interpretability is nice - but can we do better by…