SNOW blog referenced in the prior post via link-
Snowflake and Tecton help abstract infrastructure complexity for data scientists working with operational machine learning use cases where applications use machine learning to autonomously and continuously make real-time business decisions.
Perhaps this is too much in the weeds, feel free to delete; but, my take is:
Snowflake Ventures invested in Tecton because many data scientists struggle to move as quickly as needed since they lack a modular approach to machine learning which would allow for the reuse of commonly used code features. Without the modular approach Tecton makes available on Snowflake, each ML App Developer treats each new machine learning use case as custom, needing to be built from the ground up, work resulting in inefficiency and duplicated effort, resulting in many models never moving forward into production.
It sounds like Tecton platform functions as not only a central hub of the data input signals that machine learning models use to make predictions, this is a platform of features, allowing data scientists to automate, manage, and share the datasets and data pipelines required to put models into production, in a modular composable way.
Please correct me if I’m wrong; but, my understanding is that Tecton on Snowflake compliments Snowflakes functionality with Tecton being integrated into Snowflake, for Snowflakes Data sharing capabilities, increasing the use cases available for ML models; and once the ML models are created, this increases the revenue for Snowflake via Snowflake selling the needed compute that the ML applications require.
Snowflake is enabling Tecton to work and Tecton is increasing the use cases for which Snowflake can bill. I like this virtuous circle for Snowflake!