Bot flow orchestration

AI-generated summary

Genesys Cloud's Call Bot Flow action enables integration of digital bot flows into messaging systems through a drag-and-drop Architect interface, supporting both new bot creation and existing bot integration with configuration options for action naming, bot flow selection, and variable management. The feature requires CX 1-4 licensing and appropriate permissions for Architect UI View and Language Understanding access. The system enforces technical constraints including a 1000-action per turn limit, 200-turn session maximum, and five-iteration recursive task restriction. Execution tracking occurs through specific exit codes such as TriggeredByFlow and AgentRequestedByUser, while error handling mechanisms address scenarios including UnknownIntent, InvalidInputVariable, InvalidConfirmationMode, and multiple RecognitionFailure subcategories. Bot flow architecture distinguishes between core flows, which are invoked by inbound flow types at the top level, and subflows, which are invoked by other flows. Bot flow authors can transfer control between flows using Call Bot Flow actions within bot flows or Call Digital Bot Flow actions within digital bot flows. Each bot flow maintains independent configuration and limits, with Natural Language Understanding and knowledge fabric scoped individually per flow. Multiple bot flows can execute within a single interaction session, supporting voice-to-voice and digital-to-digital calling patterns. Language inheritance allows subflows to adopt the core flow's language when configured, while the core flow's Voice Processing Prompt configuration applies system-wide. Genesys Cloud provides comprehensive monitoring through the Bot Conversation Library view, Utterance History view, Execution History, Learning, and Flow Insights data for both core flows and subflows. Bot Performance, Flow Performance, Virtual Agent Performance reports, and the Optimization dashboard display data exclusively for core flows. The Test & Start Chat feature allows users to preview all bots in a chain when testing core flows, while testing subflows restricts participation to the specific bot flow and its child bots. Virtual Agent-enabled flows support bot-to-bot invocation, with conversation and wrap-up codes functioning only when the core flow is Virtual Agent-enabled. The platform provides comprehensive documentation and support resources including articles on Dialog Engine Bot Flows, Architect, bot configuration, and bot flow creation tutorials, with users able to submit feedback through multiple channels by specifying their role and reason for feedback

In Architect, bot flow authors can add a Call Bot Flow action within a bot flow to transfer control to another bot flow. Similarly, authors can use a Call Digital Bot Flow action within a digital bot flow to transfer control to another digital bot flow.

The following terminology applies to nested bot flows and digital bot flows:

Core flow: A flow that is invoked by an inbound flow type, such as a message, call, or chat. A core flow either exists at the top level of a flow hierarchy or operates as a stand-alone flow. Before the introduction of this feature, all bots functioned as core flows.
Subflow: A flow that is invoked by another flow rather than directly by an inbound flow. A subflow can be called by a core flow or by another subflow. A subflow is also referred to as a child flow.

Note: The Architect UI does not distinguish between core flows and subflows. Any bot can function as either a core flow or a subflow, depending on how it is referenced within the flow hierarchy.

When building a bot flow that calls one or more existing bot flows, consider the following limitations and guidelines:

Each bot flow maintains its own configuration and limits.
Multiple bot flows can run within a single interaction session.
Bot flows can call other bot flows of the same type – voice to voice and digital to digital.
Natural Language Understanding (NLU) is scoped to each bot flow and is not shared automatically.
Knowledge fabric & knowledge bases are scoped to each bot flow and are not shared automatically.

More considerations

Virtual Agent-enabled flowsClick to expand

Language inheritanceClick to expand

A subflow inherits the core flow’s language if the subflow is configured to support that language. Similarly, a core flow inherits the subflow’s language if the core flow is configured to support that language.

Example

Bot A and Bot B support both English and French. If Bot B switches the conversation language to French and then passes control back to Bot A, the conversation language remains French.

Voice processing promptClick to expand

Bot metrics in the Optimization dashboardClick to expand

The Optimization dashboard displays metrics only for the core bot flow. It does not display metrics for subflows invoked by the core flow.
The Bot Conversation Library view displays metrics for both the core bot flow and subflows.
The Utterance History view displays utterance details for the current flow, whether it is a core flow or a subflow. When viewing the utterance history of a core bot flow in a nested bot flow hierarchy, you can click the subflow name to open the subflow. Similarly, when viewing the utterance history of a subflow, you can click the core flow name to open the core flow.

Execution History, Learning, and Flow InsightsClick to expand

Bot Performance, Flow Performance, and Virtual Agent Performance reportsClick to expand

Test botsClick to expand

When you test a digital bot flow using the Test > Start Chat feature, the system allows you to preview all bots in the chain.

For example, when testing the core flow, you can participate in a conversation that spans all the bots involved, such as Core Bot > Sub Bot 1 > Sub Bot 2. Similarly, when testing subflows, you can participate only with the specific bot flow and the child bots, such as Sub Bot 1 > Sub Bot 2.

[NEXT] Was this article helpful?

Get user feedback about articles.

Loading...

Bot flow orchestration

More considerations

[NEXT] Was this article helpful?