We specialize in tabular data and are building a pipeline-based approach for creating and serving models.