Query Manager

The TranzAI platform simplifies the management of queries and the extraction of data series for feature extraction, data analysis and visualization.

No-code query manager

All data consolidated in the feature store can be extracted through a no-code query manager, making it easy to create data series from any data set acquired with TranzA data pipelines.

These data series can then be combined, merged or refined in the analytical components of the TranzAI platform to perform exploratory (spatial) data analysis and create dashboards.

SQL IDE

Experienced developers can completely bypass the visual query editor and directly use the SQL IDE to write custom queries that best suit their needs.

Queries visualization

You can dynamically visualize the results of your queries in data tables and progressively refine your queries to meet your analytical requirements and the constraints of the chart editor.

Queries explorer

Queries can be saved to be used by different processes and users. The TranzAI query explorer provides all the information on the different queries built by users within a project. You can build queries from existing queries to save time and refine your analysis.

Exploratory data analysis and dashboard creation are iterative processes that can be easily documented and shared by data analysts with automatic query documentation and logging.

Dynamic queries

Dynamic queries are queries whose parameters can be dynamically bound to the TranzAI frontend to build entity pages or dashboards that will use these parameters to query data according to the navigation context and user interactions.

Query management workflow

Create query from table

The no-code query editor allows you to build your SQL query using the metadata of any table referenced in your feature store.

Create query from data source

You can query raw data from a data source that has been defined in the TranzAI platform. The no-code query editor uses the metadata of the data source schema to build the associated queries.

Create query from training data set

Building queries dedicated to TD sets analysis can be very helpful for EDA and drift analysis. Queries are also built from the metadata of the TD set that are automatically created after each TD set pipeline execution.

Create custom query

If you are an SQL expert and have an in-depth knowledge of the feature store data model, you can directly write custom queries for analytical purposes or feature extraction.

If you get access to a remote data source (BigQuery, Athena, Snowflake...), you can securely store your credentials to interact with your remote data and perform analytics or data enrichment operations without the need to duplicate your data.