Are You Using Data Pipelines for Machine Learning Models?

We spend more time with folks who are generating a lot of these curated high fidelity data sets to train the models on. For example, you can't do ad hoc queries very cost effectively and efficiently against two billion data points every single day for your BI report. That's what we see a lot of people using those downstream data pipelines for. We largely talking about things that are, let me call it isolated pipelines or data comes in from some original source.