Outcomes overview

From Genesys Documentation
Jump to: navigation, search

Genesys Predictive Engagement's AI-powered outcome scoring service learns to predict your business outcomes using machine learning models that are unique to your business.

Predicting outcome scores

Genesys Predictive Engagement uses AI to predict whether a customer will achieve a particular business objective. These business objectives are called outcomes. For each outcome you define, there is a machine learning model that evaluates a user's behavior against the behavior of other users on your website to determine the user's outcome score. Your organization's models are unique to you.

Each outcome score represents the likelihood that the user will achieve the particular business outcome, based on the actions the customer has taken so far during the session or on other activities related to the customer that are included in the appropriate events (for example, geolocation).

The model updates a user's score for each outcome in real-time, and a user's scores can change as they navigate your website.

If a customer ends up talking with an agent, the agent can see the customer's outcome scores while viewing the customer's complete set of journey context data. In addition, the outcome scores can be used to trigger action maps that enhance a customer's engagement on your site.
Because our internal service handles the training and deployment of models, it is not possible to deploy customer-specific models with Genesys Predictive Engagement. Our team investigates and integrates new algorithms into the scoring service based on customer use cases.

How Genesys Predictive Engagement gathers outcome probability data

Genesys Predictive Engagement monitors all the ways your customers arrive at and interact with your website pages. For example, if you are an e-commerce site, Genesys Predictive Engagement tracks how customers navigate your site and place items in their shopping cart as they proceed to the checkout page.
  • How a visitor achieves a certain probability score is unique to your business and website.
  • Outcome scores and the associated data science are scored in a GDPR-compliant manner. Predictive Engagement's data scientists work exclusively on anonymized GDPR-compliant data.
For more information, see About the data we track.

Start training your models

Each of your machine learning models must be trained before it can make predictions. In order to start the training, you must:

  1. Create an outcome.
  2. Have users actively using your tracked website. To verify user activity, use Live Now.
After you complete these steps, the model training automatically begins within 24 hours. The training process is completed and the model is ready to work before the start of the next business day.
Initially, you will see only a green bar with a check next to it in the Outcome Scores section. This indicates that the outcome condition has been met. You will see outcome scores after your model has been trained.

Ongoing training

The model training process is fully automated: you do not need a data scientist to start, monitor, or maintain the training. Your model is automatically retrained nightly using the last 30 days of your user data. In addition, your model is periodically evaluated and retrained on fresh data.  

While the model is undergoing retraining, outcome scores are predicted using the previously trained version of the model. During the training process, users will be able to continue working as normal.

Your newly trained model is tested to ensure it performs better than the previous version in terms of its precision and recall. If the new model does not work at least as well as the previous model, the previous model is reinstated until more data is gathered. Historical models are not retained.
  • When you add a new outcome, your model training automatically accounts for it.
  • Automated model training and predictions are performed on the original customer journey events, which are not anonymized and will generally contain PII.

Improve predictions

In general, the longer your models run and the more data they evaluate, the better their predictions will be.

The best way to improve a model's predictions is to increase the number of achieved outcomes. In general, your dataset should contain several hundred positive examples in order for your model to be adequately trained to make reliable predictions.

Other factors can affect the accuracy of your model's performance, including:

  • The total number of customer journeys recorded
  • The frequency that an outcome occurs in the data
  • The richness of events produced by the customer
    You can exclude IP addresses to prevent internally generated events from influencing your models.

Predict based on custom events

Customers can provide predictions generated by other predictive systems as "custom events." Custom events can be consumed by the Outcome Score Service and used to train models and generate outcome predictions based on the features produced by an external system. This is another example of how a generalized AI system enables hyper-personalization.

**We need to put this in plain English, but I want clarification before I begin...


Is this current-state, or tied to AI-23?