Jump to: navigation, search


All Genesys software is © Copyright 2016–2018 Genesys Cloud Services, Inc. All rights reserved.

Complete information about Genesys proprietary intellectual property, including copyrights, can be found here.


Genesys and the Genesys logo are registered trademarks of Genesys Cloud Services, Inc. in the U.S.A. and other countries. Complete information about Genesys proprietary intellectual property, including all trademarks, can be found here.

All other trademarks are the property of their respective owners.

Third Party Software

Genesys follows applicable third-party redistribution policies to the extent that Genesys solutions utilize third-party functionality. Please contact your customer care representative if you have any questions. The following list describes specific third-party code and functionality for this product:

  • This product utilizes MongoDB 3.0.5. The corresponding source code is available here: https://www.mongodb.org/.
  • NOTICE OF RESTRICTED RIGHTS FOR ORACLE PRODUCTS LICENSED TO THE US GOVERNMENT Oracle Programs delivered to the United States government subject to the DOD FAR Supplement are 'commercial computer software' and use, duplication, and disclosure of the programs, including documentation, shall be subject to the licensing restrictions set forth in the applicable license agreement therefor. Otherwise, Oracle programs delivered subject to the Federal Acquisition Regulations are 'restricted computer software' and use, duplication, and disclosure of the programs, including documentation, shall be subject to the restrictions in FAR 52.227-19, Commercial Computer Software-Restricted Rights (June 1987). Oracle USA, Inc., 500 Oracle Parkway, Redwood City, CA 94065.

New in AICS Strategy Subroutines (07/19/2019)

  • The NGINX container is no longer included in AICS. Previously, NGINX was provided as a convenience for use in internal and/or pre-production environments.
  • This release enables you to log scoring details that you can then use to monitor and better understand the scoring process and outcomes. This release also includes scripts to clean up unneeded score logs from MongoDB. To how to turn on score logging, for information about what is logged, and for the cleanup script, see View the Scoring Logs.
  • New API endpoints have been added, enabling you to use the MinIO container to upload Agent Profile and Customer Profile data. Previously, only Dataset data used the MinIO container, which provides a performance improvement over the data_upload worker alone. For details about these new endpoints, see the Predictive Routing API Reference, which you can access from a link on this page. (Requires a password for access. Please contact your Genesys representative if you need to view this document.)
  • The GPR API now returns a file_path parameter in the response message when you request a presigned URL, which replaces the uploaded file name. You must now pass this file_path parameter in requests to create Datasets or the Agent Profile or Customer Profile instead of the uploaded file name, used in previous releases. For details, see the Predictive Routing API Reference, which you can access from a link on this page. (Requires a password for access. Please contact your Genesys representative if you need to view this document.)
  • The GPR web application has enhanced security by logging out inactive users.
  • GPR now accepts only CSV and zipped CSV files for upload to Datasets, the Agent Profiles, and the Customer Profile. JSON file uploads are no longer supported.
  • The numeric datatype now replaces both float and integer datatypes in Agent and Customer Profiles. This change resolves confusion about when to use float and integer datatypes.
    This is a breaking change from previous releases of AICS and Agent State Connector (ASC). If you upgrade either component you MUST also upgrade the other. The following versions are compatible:
    • AICS and higher + ASC and higher.
    • AICS and lower + ASC and lower.
  • Improvement in the queries used to run the Feature Analysis Report improve the speed and reliability with which these reports are generated.
  • The Deployment and Operations Guide now contains complete instructions for configuring HTTPS connections among all GPR components: Configure GPR to Use HTTPS
  • You can now optionally run ASC without a connection to Stat Server. ASC automatically detects whether there is a Stat Server Application in the ASC Application Connections tab. If so, ASC connects as in previous releases. If not, ASC does not connect to Stat Server and does not send agent login status updates to the GPR platform. In this case, AICS receives agent availability information from Universal Routing Server (URS).
  • ASC now validates the values you enter in the include-skills and include-groups configuration options. If ASC detects a skill name or Agent Group specified in these options that does not exist in Configuration Server, ASC triggers a Standard-level log message (Message Server log event number: 60401).
  • A number of subroutines have been introduced or modified. They perform the following functions:
    • The new GPRIxnSetup subroutine initializes all the GPR KVPs with default values, setting with gpmMode to off and gpmResult to 15 (Predictive Routing is turned off or not used for this interaction).
    • The GPRIxnCompleted subroutine has been updated to include the correct value of gpmScoreAboveMedian KVP by comparing the selected agent score with the returned median score.
    • The ActivatePredictiveRouting_v3 and GPRIxnCleanup subroutines have been enhanced to add connection_ids as a URL parameter in the score request and the score_log request, for easier troubleshooting in URS logs.
    • The GPRIxnCleanup subroutine now performs score log and UserEvent (KVP) distribution (previously done in the GPRIxnCompleted subroutine). This subroutine can now correctly identify abandoned interactions and interactions in which GPR was unable to route the interaction and add this information to the score log and the Genesys Info Mart gpmResult KVP, with the values 13 - Call Abandoned and 14 - Call Routing Failed.
  • The URS Strategy Subroutines now send the following new KVPs, which are stored in the Genesys Info Mart database and are available for reporting:
    • gpmAdjustedAgentScore - The final agent score used to route the associated interaction to the selected agent. This score is calculated from the gpmAgentScore combined with any agent occupancy factor.
    • gpmDefaultAgentScore - This default agent score for the associated interaction. The value is the outcome, for this interaction, of the setting specified in the default-agent-score configuration option.
    • gpmDefaultScoredAgents - The number of agents with default scores assigned for an interaction.
    • gpmDefaultScoreUsed - Indicates whether the agent score for the associated interaction is taken from the scoring response returned by GPR or the value of the default-agent-score configuration option.
    • gpmGlobalScoreCount - The number of agents scored using the Global model. This value is the content of the global_scores_count field returned by AICS in the scoring response.
    • gpmInitialScoreThreshold - The initial threshold value used for the interaction, taken from the value set in the score-base-threshold configuration option.
    • gpmFinalScoreThreshold - The final threshold value used to route the associated interaction to the selected agent. The routing strategy calculates the value from the configured score threshold combined with values resulting from any agent holdout options.
    • gpmScoreAboveMedian - Indicates whether the score for the selected agent was better than the median score for the target group.
    • gpmSuitableAgentsCount - The number of agents who had scores greater than or equal to the initial threshold value when the scoring response was received.
    • gpmPredictorType (Reserved for future use)
    • gpmRoutingMethod (Reserved for future use)
    For additional information see the following sources:
  • The gpmResult KVP now includes four new values:
    • 12 - Reserved for future use
    • 13 - Call Abandoned
    • 14 - Call Routing Failed
    • 15 - Predictive Routing is turned off or not used for this interaction
  • The value off was added to the valid values for the gpmMode KVP.
  • The gpmWaitTime value is now calculated using START_TS rather than gpm-ixn-timestamp.
  • The GetActionFilters subroutine was enhanced to identify the list of agents matching the target skill group along with the configured login status expression. This information is also reported in the action filters of the scoring request. This functionality is invoked only when the use-action-filters configuration option is set to false.
  • Two new documentation pages, Routing Scenarios Using GPR and How Does GPR Score Agents? provide detailed discussions of those aspects of GPR functionality.

New in Release (03/26/2019)

  • This release includes an updated and improved version of the Predictive Routing API Reference, which you can access from a link on this page. In particular, there are now cURL request examples for each endpoint.
  • AI Core Services now requires Docker version 18.09.2, which addresses important security issues. See the Release Notes for AI Core Services for important information about the reason for this change and for Docker deployment information.
  • You can now convert a regular account into an LDAP account. A toggle with label LDAP has been added to the Settings > Account update window. To enable LDAP authentication, enter your LDAP credentials and save changes. After that, you must also convert the user accounts for those who should use LDAP authentication. User configuration is done in the Settings > User Management window.
  • This release includes the following improvements to the user interfaces in the GPR web application:
    • A new navigation panel provides a tree view of all Datasets, Predictors, and Models configured for the current Tenant. This tree-view pane is available from the Settings > Datasets and Settings > Predictors windows. Each item in the tree view links to the specified object, enabling easy access to the entire hierarchy of Datasets, Predictors, and Models.
      NOTE: Composite Predictors, which can be built on data from multiple Datasets, are not shown in this tree-view pane.
    • For simplified navigation, breadcrumb links now appear at the top of windows in the GPR web application if you have drilled-down past a top-level window.

New in Release (12/21/2018)

  • You can now upload Dataset, Agent Profile, and Customer Profile data to Genesys Predictive Routing (GPR) from CSV files that use certain legacy encodings (listed below). By default, GPR always assumes the CSV files are encoded with UTF-8. This change applies to uploads using both the GPR web application and the GPR API. The following encodings are supported:
    • UTF-8
    • Shift JIS
    All data returned from GPR uses UTF-8 encoding.
  • GPR has optimized how cardinalities are stored. Cardinalities are now written into a dedicated database collection, so that the cardinalities for each field are stored in their own document. Previously, the cardinalities were stored along with the schema data. With high-cardinality features, this could lead to performance degradation due to additional conversions needed to extract the cardinality data.
  • The schema management workflow for Agent and Customer Profiles has been simplified and streamlined. The Discovered Fields tab has been removed and cardinality counts have been added to the schema view. This change ensures GPR always presents up-to-date Profile information. The schema tab always presented updated information, if available, but the Discovered Fields tab display was generated only once and did not reflect changes to the Profile schema.
  • The explanation for how to configure and interpret the Agent Variance Report has been clarified and expanded.
  • AI Core Services (AICS) now supports deployment in an environment running on a Kubernetes cluster.
  • AICS supports Security Enhanced Linux (SELinux) on CentOS 7.
  • If the ASC configuration contains non-empty values for the new filter-by-skills and/or filter-by-groups configuration options, ASC subscribes to Stat Server for agent statistics only for the agents included in the specified Agent Groups or those satisfying the configured skill expression. If both options are configured, the agents are subscribed for statistics if they either satisfy the skill expression specified in the filter-by-skills option or are included in one of the Agent Groups specified in the filter-by-groups option. This functionality enables you to limit the number of agents monitored by GPR or to use GPR in environments where multiple Stat Servers are deployed to monitor different groups of agents.
  • Agent State Connector (ASC) now supports email interactions, as well as voice.

New in Release (10/25/2018)

  • Dataset handling has been made significantly faster by means of the following improvements:
    • For the initial data upload, this release introduces the Minio container.
    • The Dataset import to MongoDB now uses a multithreaded process.
  • The NGINX container has been removed from AICS. NGINX is an optional load balancer that had been provided only for use only in test environments.
  • The Sizing Guide for Genesys Predictive Routing (GPR) has been entirely reworked and expanded.
  • The explanation for how to use Composite Predictors has been revised and clarified.
  • The GPR web application and GPR API now use the same process to create Agent and Customer Profile schemas. In addition, the instructions for Configuring Agent Profiles have been revised and expanded.
  • The LOG_LEVEL environment variable has been added to the tango.env configuration file. By default, it is set to INFO, which is a minimal logging level, adequate for most circumstances.
  • This release upgrades AICS to Python 3.6 from Python 2.7.
  • AICS now performs automatic cleanup processes which should maintain an adequate amount of free disk space.
  • Memory handling for MongoDB was improved In this release.
  • The Lift Estimation report has been improved, adding Export functionality, the ability to toggle between graph and table displays, and showing the Aggregated view as the first tab listed.
  • This release includes a number of changes to the look and feel of the GPR web application interface, as well as multiple user experience improvements to provide more intuitive workflows and better presentation of information. For a complete list, refer to the AI Core Services Release Note for release
  • This release provides a number of improvements and additions to the GPR API, including the ability to sync and accept Datasets, check job status and support for nesting dictionary fields. For a complete list, refer to the AI Core Services Release Note for release and the Predictive Routing API Reference (Requires a password for access. Please contact your Genesys representative if you need to view this document.)

New in Release (08/23/2018)

  • The Quality column in the Models list table on the Model configuration window now includes a new metric, Local models. The metric displays the number of local models generated for agents in the dataset on which the predictor is built.
  • AI Core Services (AICS) has improved handling of UTF-8 characters. Data ingestion, model training, and analysis reports are all correctly processed for data containing non-ASCII UTF-8 characters.
  • The Genesys Predictive Routing (GPR) API now enables you to run Lift Estimation reports. The API returns a JSON response containing the Lift Estimation results. The resulting report is also automatically available for view from the GPR web application.
  • The GPR API now enables you to run Feature Analysis reports. The API returns a JSON response containing a list of features ordered by weight--that is, by the strength of the impact that feature has on the value of the target metric. The resulting report is also automatically available for view from the GPR web application.
  • The Lift Estimation report now uses the scoring expression configured for the predictor (if any) to decide whether the target metric should be minimized or maximized.
  • You can now configure Agent State Connector (ASC) to monitor the StatAgentOccupancy Stat Server statistic.
  • You can now configure ASC to monitor a subset of the total list of agent groups present in agent profiles.
  • You can now choose to have ASC ignore the following unsupported ASCII characters: [Space], -, <, >.
  • You can now configure ASC to monitor a subset of the total list of skills present in agent profiles.
  • ASC now supports a connection to Stat Server running in single-server mode, without a backup.

New in Release (07/13/2018)

  • This release includes a number of additions to the Predictive Routing API:
    • You can now generate and purge predictor data.
    • You can now create a new predictor by copying an existing one.
    • You can now use GET commands to retrieve dataset and predictor details.
    • The way Predictive Routing recomputes cardinalities when you append data to Agent or Customer Profiles using the API has been improved.
    • You can now retrieve information on the currently deployed platform using the new version endpoint.
  • This release includes the following new supported platforms:
    • Mongo DB 3.6 (requires a special upgrade procedure; see the AI Core Services Release Note for details)
    • Oracle Linux 7.3
  • You can now configure parameters to control password-related behavior such as how often users must change them, blocking users after a specified number of login attempts, and adding a custom message when users are blocked.
  • The audit trail functionality has been improved, to record additional actions and provide the ability to specify how long audit trail records are kept. All actions related to logins, object modification/creation/deletion, and so on, whether performed using the GPR application or the API, are logged.
  • You can configure the Predictive Routing application to display custom messages on the login screen.
  • You can now upload data (agent, customer, and dataset) using zip-archived .csv files.
  • Predictive Routing now correctly recognizes columns with any combination of the following Boolean values: y/n, Y/N, Yes/No. Previously, only columns with true/false and 0/1 values were discovered as Booleans. The identification is case insensitive.

New in Release (05/11/2018)

  • Genesys Predictive Routing complies with GDPR requirements for handling sensitive customer information. The Predictive Routing API has been expanded to provide Read and Delete functions, enabling you to locate and remove user data. For details, refer to Handling Personally Identifiable Information in Compliance with GDPR Requirements in the Genesys Predictive Routing Deployment and Operations Guide and to the Predictive Routing API Reference (access requires a password; contact your Genesys representative for assistance).
  • The Lift Estimation report now offers Advanced Group By functionality, which provides more flexibility in customizing the report. For details, refer to the Lift Estimation Report Overview in the Genesys Predictive Routing Help.

New in Release (03/28/2018)

  • Journey Optimization Platform (JOP) was renamed to AI Core Services (AICS).
  • When generating the Lift Estimation report, Predictive Routing now provides the option to produce a report for each unique value for a selected column (feature). Previously, any feature with a cardinality of more than 20 was excluded, which meant that you could not produce reports with a granularity higher than 20 unique features.
  • The agent pool for lift estimation is now constructed on a per-day basis for the interactions in the dataset. Previously, you might have observed a negative lift for higher agent availability or an unexpectedly high lift for low agent availability due to overcorrection caused by a mismatch between the input sample size and the actual sizes encountered through daily simulation.
  • Predictive Routing now supports LDAP authentication when logging in.

New in Release (03/05/2018)

  • You can now enable Predictive Routing to look up updated values for certain agent attributes, based on customer or interaction attributes during a scoring request. For instance, you can look up agent performance by virtual queue, enabling you to evaluate the agent’s previous performance when handling interactions from that queue. This avoids comparing agent performance for a specific queue against other agents who handle interactions from a different mixture of virtual queues.
  • You can now view an entire Agent Profile or Customer Profile record from the Agents Details and Customers Details tabs or an entire record on the Datasets Details tab. Click a single record to open a new window containing a table with all the related key-value pairs.
  • Agent State Connector now supports connection to a secured Configuration Server port and TLS 1.2 connections to Stat Server.
  • You can now configure Agent State Connector to automatically create an Agent Profile schema, if none exists, or to verify the existing schema.
  • You can now have Agent State Connector collect call connId data from Stat Server and write it to the Agent Profile schema for use in Predictive Routing.
  • Agent State Connector is now supported on Windows. For exact versions supported, see the Genesys Supported Operating Environment Reference Guide.
  • Predictive Routing now supports datasets of up to 250 columns for predictor data generation, model training, and analysis.
  • Model training speed has been considerably improved.
  • Predictive Routing now provides progress indicators when loading predictor data and generating predictors. The progress indicators show the percent complete and the number of data rows already loaded.
  • The maximum supported cardinality for the Group By functionality in the Lift Estimation report has been increased to 20. All features with cardinalities between 1 and 20 are now available in the Group By selection menu.
  • You can now enter a maximum value of 500 simulations in the Lift Estimation analysis report settings. This prevents you from entering numbers too large to efficiently analyze and which can lead to an out-of-memory situation. The Number of Simulations field accepts any value larger than 0 and less than or equal to 500.
  • Predictive Routing now provides a text search field for use when selecting attributes for analysis.

New in Release (01/05/2018)

  • The user interface and the documentation have been updated to reflect the product name change from Genesys Predictive Matching to Genesys Predictive Routing.
For now, the following Genesys Predictive Routing components retain "Predictive Matching" as a part of the component name: Predictive Matching - Agent State Connector, Predictive Matching - Composer Strategy Subroutines, and Predictive Matching - URS Strategy Subroutines.

New in Release (12/22/2017)

  • The product name has changed from Genesys Predictive Matching to Genesys Predictive Routing. This change is not yet reflected in the application interface or in the documentation.
  • Genesys Predictive Routing now supports both single-site and multi-site HA architectures.
  • Genesys Predictive Routing now supports historical reporting, provided by the Genesys Reporting solution. The following reports are available in Genesys Interactive Insights/GCXI: Predictive Routing AB Testing Report, Predictive Routing Agent Occupancy Report, Predictive Routing Detail Report, Predictive Routing Operational Report, and Predictive Routing Queue Statistics Report.
  • Two new real-time reporting templates are available for use in Pulse dashboards: Agent Group KPIs by Predictive Model and Queue KPIs by Predictive Model.
  • Two new analysis reports have been added to the Genesys Predictive Routing application: Agent Variance and Lift Estimation.
  • The Model creation interface now includes additional model quality and agent coverage reporting.
  • The Feature Analysis report, the model creation and training functionality, and the dataset import functionality have been improved to handle large datasets.
  • You can now combine simple predictors to create composite predictors.
  • Health checks and monitoring have been improved for both Journey Optimization Platform (JOP) and Agent State Connector (ASC). ASC now enables you to set alarms if there are persistent connection issues with Configuration Server or Stat Server.

New in Release (09/26/2017)

  • You can now deploy the Journey Optimization Platform (JOP) in Docker containers.
  • This release includes updates to the Predictive Matching web interface for improved usability and rebranding.
  • This release includes context-sensitive Help.
  • You can now update and retrain models that have not yet been activated. You can also make changes to activated models by cloning them, editing the parameters, then activating the new model in place of the old one.
  • Agent State Connector and the Strategy Subroutines components can be deployed in a high availability configuration.
  • Predictive Matching now supports HTTPS.
  • Predictive Matching now supports TLS 1.2 encryption. Support for TLS 1.1 has been discontinued.
  • Routing using Predictive Matching can now take agent occupancy into account when selecting the best target.
  • The workflow for creating Predictors has been made more logical and straightforward.
  • Users can now reset their passwords from the Predictive Matching web interface.

New in Release (06/27/2017)

  • Predictive Matching now provides REST APIs for scoring, agent profile updates, and customer profile updates.
  • Predictive Matching now supports strategies created in Composer and processed by Orchestration Server (ORS). These strategies utilize common Universal Routing Server (URS) subroutines to store scores returned from the scoring server and to set callback functions in URS.
  • Schema modification has been extended to enable manual creation of fields not included in an imported dataset. This extended functionality also enables discovery of additional fields by uploading further data and thereby extending the schema.

New in Release (03/28/2017)

  • Extended datasets functionality now includes built-in analysis capabilities to make data exploration and feature analysis more straightforward without requiring customers to first build a predictive model.
  • Customer profile data can now be loaded to the platform by means of a REST-API and joined at run time for scoring. This simplifies the integration requirements for deploying Predictive Matching, requiring less modification to existing routing strategies or run time CRM integrations.
  • Predictive Matching now enables logging of routing decisions, required for accurate A/B testing, to JOP rather than Genesys Info Mart. This simplifies Predictive Matching deployment, by removing the need to make changes to Interaction Concentrator and Genesys Info Mart to support Predictive Matching.
  • The new Predictive Matching Help now opens when you click the Help link in the Predictive Matching interface.

New in Release

  • Improvements to the analytics and reporting functionality:
    • Reports can now indicate whether a predictive score was generated (that is, A/B testing), and whether it was interleaved, or time-divided.
    • The range of visualization on the Reporting Dashboard page has been improved.
    • Predictive Routing can now perform analysis and data discovery on factors driving the KPI that is being optimized.
  • You can now upload data sets in CVS format, enabling you to have Predictive Routing analyze the data, define predictors based on it, and report on it. You can use these data sets for model training and testing, and you can calculate statistics for correlation and cardinality from them.
  • Self-service predictor management and model creation. Note the following properties of predictors and models:
    • You can have multiple models built from one related data set.
    • You can only use a predictor to optimize a single metric (that is, a column in the data set); each model under the same predictor optimizes the same metric.
    • You can use a subset of features from the data set to define a predictor.
    • A predictor can be based on a subset of data (such as a time range, or a subset created by filtering data set column values).
    • Once you define a predictor, you can append new data to its underlying data set.
    • A predictor can use any source of data matching the source data set schema to retrain and update models.

New in Release

  • Routing subroutine support to score agents in agent-surplus mode, where there are more agents in ready status than interactions requiring agent handling. Agents can be scored on various criteria you configure so that interactions go to the most suitable agent.
  • Routing subroutine support to score agents in customer-surplus mode, where more customer interactions are waiting in queue than there are available agents. Interactions are scored based on criteria you define so that customers you consider highest priority are handled first.
  • Logging, monitoring, and alarm capabilities.
  • Ability to combine various criteria for scoring agents or interactions for suitability, leading to a more nuanced matching of agents and interactions.
  • Fault­ Tolerance. If Predictive Routing is unavailable, or if a Predictive Routing strategy subroutine takes more than a specified amount of time to process an interaction, the routing strategy defaults back to standard behavior.

Genesys Predictive Routing

Genesys Predictive Routing draws on accumulated agent, customer, and interaction data, enabling you to analyze omnichannel interactions and outcomes and generate models to predict outcomes. From this analysis, combined with machine learning, you can determine the best possible match between waiting interactions and available agents, and then route the interactions accordingly.

In addition, you can report on the predicted versus actual outcomes. The actual outcome is also used to further train the machine-learning model, improving the accuracy of predicted outcomes between similar customer profiles and agent profiles.

In addition, Predictive Routing includes an open REST-API for scoring and for providing feedback, enabling continuous or periodic automated improvement of models.

In releases prior to, AI Core Services (AICS) is known as Journey Optimization Platform (JOP).

What's New

Genesys Predictive Routing is part of 9.x, which can include component releases from 9.1.x, 9.0.x, and 8.5.x code streams. Use the table below to check which component releases are part of 9.x.

All 9.x products 9.x Genesys Predictive Routing Release Notes
Product Component Latest Starting

Genesys Predictive Routing

AI Core Services

Genesys Mobile Services

Predictive Routing - Agent State Connector

Predictive Routing - Composer Strategy Subroutines

Predictive Routing - URS Strategy Subroutines


Deployment and Operations Guide

Provides Genesys Predictive Routing system requirements, sizing, deployment, scaling, logging, starting and stopping, and monitoring information. It also explains how to have Predictive Routing supply data for Genesys historical and real-time reporting and how to integrate the Predictive Routing subroutines into your Genesys Routing environment.

Help File

Explains how to use the Genesys Predictive Routing user interface, including explanations of how to upload and update your data, create Predictors and Models, and how to run reports analyzing your data to create the most effective Predictors and Models.

==== Predictive Routing API Reference

The API Reference for Genesys Predictive Routing.

New Features and Modifications

Records new features and functionality in all 9.0.0 releases, with links to the relevant documentation.

Related Products

More Release Information

This page was last edited on November 18, 2021, at 20:44.