Feedback & Training
Giving Feedback is the first step of the learning process. Through feedback we collect human decisions on events, without those decision being explained by the analyst. In other words, we will be collecting the final result of the investigation and the decision points, without being directed as to why a certain decision was made, or which of the decision points mattered more in reaching it.
Training is the second step, where the information collected through feedback is being assimilated by the AI models. Each training round will add more and more information to the model's knowledge base, making it more and more aware of the context and the data patterns.
The models are thoroughly vetted through KPIs that give insights into their performance. Any anomaly into the training process, such as conflicting data or wrong feedback given over time, will reflect into metrics that can trigger notifications, pipelines, or alerts.
For additional details, check the User Guide -> Feedback and Training section