TileDB-ML
Store ML models alongside direct access to your multi-modal datasets for training and prediction. Support for TensorFlow-Keras, PyTorch and scikit-learn.
SQL
Access TileDB arrays using a range of SQL engines, including MariaDB, Presto, and Trino.
BI and Analytics
Connect to traditional enterprise BI and analytics tools using the type 4 JDBC driver for TileDB Cloud, and power visualizations using the TileDB-Tableau-Connector. Directly support serverless SQL queries within Python on TileDB Cloud using the Python DB API.
Interoperability
In addition to TileDB-maintained integrations, TileDB arrays are designed to work seamlessly with pandas, NumPy, and R data frames. Wherever possible, TileDB uses Apache Arrow for zero-copy integration.