Data storage

At TranzAI level, storage is decoupled from compute. By using a cloud object store, TranzAI users can scale their storage needs as their data science projects grow. They automatically benefit from the best cost optimization options implemented at TranzAI level (compression, column storage, type optimization, etc.).

In the TranzAI platform, all data processing is initiated in a scale-up and scale-down mode. Computational costs are limited to the volume of data processed and analyzed by the data science team for their projects, independent of storage costs.

TranzAI data storage leverages cloud-based horizontal scalability while providing specific GIS and spatial features.

Parquet data

When we need to consolidate data from remote sources and store them locally in your TranzAI feature store, we rely on the Parquet format to optimize costs and performance.

Spatial backend

The TranzAI platform includes a spatial backend dedicated to the processing of spatial data using dedicated spatial functions and operators.

According to the origin of your spatial data and the environment used to process and extract spatial features, The TranzAI platform will automatically manage the conversions between different formats (GeoJSON, WKT) and coordinate reference systems.