Your business changes over time. Your business changes may (or may not) impact the performance of machine learning pipelines which are dependent on data generated by the business. Cortex users can now assess how the performance of their machine learning pipelines are changing over time by specifying training dates in the past during pipeline creation.

And it’s possible with only one additional click in the no-code interface.

Ensuring High Performance and Stable Machine Learning Pipelines

Prior to deploying a machine learning pipeline it’s important to have confidence that the pipeline’s performance will be stable over time. One way to gather confidence in your new pipeline is to train the pipeline using only data up to a specified past date and validate pipeline performance based on data up to that date. The rationale being, if your pipeline was relatively stable over multiple dates in the past, it’s reasonable to assume it will be stable in the future as well.

Cortex now makes this type of time-travel analysis easy. Simply select a date in the past when creating your pipeline and Cortex will train using only data up that past date.

Cortex snapshot highlighting the ability to specify a past date when creating a One-Time pipeline for testing purposes.

How Partners are using this Feature

Several Cortex partners are building pipelines using data up to specific months – i.e. data up to Jan, Feb, March, April, etc. and then ensuring that each of these pipelines is yielding reliable and consistent performance over time. Checks include the broad performance of the pipeline but also more detailed analysis of the conversion rates which Cortex provides after pipeline creation. These checks give the business more confidence that over time small cyclical or other changes in the business won’t adversely impact pipeline performance.

If you are interested in learning more, please let us know, we would be happy to walk through how other customers are using this feature to help build better and more reliable machine learning pipelines.  We look forward to hearing from you at

Want to Learn More?

Schedule a demo and talk to a product specialist about how Vidora’s machine learning pipelines can speed up your ML deployment and ultimately save you money.