Open ChenpiDog opened 2 years ago
@lumianph any idea?
Data is very important for machine learning, but the original data is often not easy to analyze and process, so it needs to be standardized. The data warehouse has assumed this responsibility. Most companies' data applications are built on the data warehouse. As a machine learning database, OpenMLDB can consider expanding to the data warehouse direction, such as docking with upper BI applications, real-time ETL, etc.
Data is very important for machine learning, but the original data is often not easy to analyze and process, so it needs to be standardized. The data warehouse has assumed this responsibility. Most companies' data applications are built on the data warehouse. As a machine learning database, OpenMLDB can consider expanding to the data warehouse direction, such as docking with upper BI applications, real-time ETL, etc.
@ChenpiDog thank you for your suggestions. Yes, this looks like promising, especially like real-time ETL, seems OpenMLDB will be very competitive. Do let us know if you have any concrete suggestions.
Data is very important for machine learning, but the original data is often not easy to analyze and process, so it needs to be standardized. The data warehouse has assumed this responsibility. Most companies' data applications are built on the data warehouse. As a machine learning database, OpenMLDB can consider expanding to the data warehouse direction, such as docking with upper BI applications, real-time ETL, etc.
@ChenpiDog thank you for your suggestions. Yes, this looks like promising, especially like real-time ETL, seems OpenMLDB will be very competitive. Do let us know if you have any concrete suggestions.
@lumianph Integrated data warehouse and machine learning database: 1) MPP distributed architecture based on Shared Storage is better; 2) Compatible with Postgre or MySQL connection protocols is of great significance for embracing the data ecology; 3) Real time data processing, including real-time ETL and real-time machine learning, is very competitive.
SQL is everything. Users only need to deal with SQL, and the rest should be handed over to OpenMLDB.
Can OpenMLDB be used as a real-time data warehouse? Is there any plan?