sqlflow

Why we included this project

SQLFlow is a good fit for teams that live in SQL and would rather not drag everyone into Python notebooks just to try machine learning. It extends ordinary SQL grammar with statements for training, prediction, model evaluation, and model explanation, so the same query language your data engineers already use can start driving real model work. Under the hood it compiles those extended SQL programs into Argo workflows that run on Kubernetes, which handles scheduling and distributed execution for you. It works with MySQL, MariaDB, TiDB, and Hive, and hands the actual model training to TensorFlow, Keras, or XGBoost, so it slots into stacks most teams already run. If you are looking for a way to lower the barrier between your analytics team and applied machine learning, this is a concrete path worth a close look.