Closed Make42 closed 2 years ago
Yes, that was the vision for MLDB. Since it was acquired in 2017 the system has stagnated. It's the base for research into the architecture of ML, but I would not use it for anything apart from design inspiration now.
Because of the "database" in the name, I first thought that mldb is something like a feature store. After reading a bit more it seems to me, that - instead - it is a full out-of-the-box, batteries-included type of system, where I am supposed to use the provided ML implementations (even though I can write plug-ins for the system). Which impression is correct?
My current understand is this: I use the mldb-SQL queries to define a DAG (directed acyclic graph) of data processing (including ML), the input data and the output REST ports. Is this right?