We’re introducing customer-controlled training for Classification Intelligence (CI), giving you more say over the records used to train your CI models. By hand-selecting training data, you can move faster and get more reliable results.
This update improves both model training, helping teams have advanced control to build CI models they can trust.
Here’s what’s new:
Faster, more deliberate model training: Add and remove records directly from the training set to make sure models are trained on the most relevant data.
- Add records directly to the training set
- Remove records that don’t represent the record category accurately enough
Actions that support real CI workflows: Manage training sets at scale and refine models over time without extra operational effort.
- Easily retrain or clear training sets
- Toggle to disable automatic training set regeneration for more control
With CI retraining control, you can shape how your models learn, leading to better accuracy and stronger confidence across teams.
This enhancement is available to all customers.
Want to learn more? Take a look at our help documentation or reach out to your Customer Success Manager.