Superduper: Integrate AI models and machine learning workflows with your database to implement custom AI applications, without moving your data. Including streaming inference, scalable model hosting, training and vector search.
We should have separate processes writing model outputs, while other processes are loading data and predicting.
This is particularly obvious when the inputs and outputs are large (order of magnitude of work in I/O comparable to prediction).
We should have separate processes writing model outputs, while other processes are loading data and predicting. This is particularly obvious when the inputs and outputs are large (order of magnitude of work in I/O comparable to prediction).