Teradata Warehouse Miner uses database connection session pooling so that a persistent connection to the database is not required. When an analysis is executed through the Teradata Warehouse Miner user interface, it gets a connection from the pool, generates the SQL appropriate for the analysis, and dynamically executes it through the same connection.
An entire project can be sequentially executed and/or individual analyses. Because the analyses are executed on another thread, other analyses within the project can be viewed, changed or executed. There is no theoretical limit to the number of analyses that can be executed at one time, although each one results in a different session in Teradata; care should be taken not to execute too many at once.