Time series is the most underrated specialization in ML and I will die on this hill.

Think about how much of the world is sequential data: stock prices, server metrics, patient vitals, weather patterns, energy consumption, IoT sensor readings, factory equipment telemetry.

Yet when I ask ML engineers what they specialize in, it's always NLP or computer vision. Almost nobody says time series.

That supply-demand mismatch means time series specialists get hired fast and paid well.

The tooling has gotten incredible too. Google's TimesFM is a foundation model for time series — essentially "GPT for sequential data." TSMixer applies transformer ideas. NeuralProphet extends Prophet with deep learning. PatchTST takes a clever patch-based approach.

If you're looking for a specialization that's in high demand, has low competition, and applies to virtually every industry — time series forecasting is sitting right there waiting for you.

I genuinely don't understand why more people aren't jumping on this.