Configurable conversation flows
Define prompts, recognition behavior, transitions, and conversation structure in configuration rather than scattering them through application code.
SPEECHTILES ANDAMENTO
Design configurable interactions, coordinate speech and media, and connect every conversation to your application's own data, decisions, and interface.
THE FRAMEWORK
Andamento coordinates interaction behavior while your application remains responsible for its business logic and user experience.
Define prompts, recognition behavior, transitions, and conversation structure in configuration rather than scattering them through application code.
Coordinate supported recognition and spoken responses as part of an interaction—not as disconnected application calls.
Bring media operations into the same conversational execution model, with behavior determined by the configured interaction.
Supply and update values that influence prompts, decisions, and how a conversation proceeds.
Respond to framework events, perform application work, and influence conversation execution through supported interfaces.
Coordinate conversation behavior around supported audio and system interruptions. Use configurable pause, resume, repeat, or continuation behavior where supported, with application callbacks and platform-specific handling.
The developer documentation describes exact API behavior, supported combinations, and platform-specific considerations.
INTEGRATION IN PRACTICE
Andamento is not a separate assistant competing with your app. It is a framework your app uses to execute conversational behavior.
Conversation definitions describe the interaction. The runtime coordinates execution. Your app handles its own data, operations, and user interface.
Explore the architecture →Configure prompts, recognition behavior, and transitions.
Provide required resources and connect callbacks and runtime data using the documented SDK sequence.
Andamento coordinates the interaction; your application responds to events and makes business decisions.
Update conversation data and application UI as appropriate for the interaction.
ANDAMENTO + GENERATIVE AI
Generative AI and Andamento complement one another. An application can ask an AI service to interpret an open-ended response or draft content, evaluate that output, and supply appropriate results to a configured conversation.
The application selects the AI service, handles validation and business rules, and decides what information enters the conversation.
Illustrative integration pattern. This does not imply a built-in LLM connector or AI-specific SDK API.
EXAMPLE · SERVICE REQUEST
Conversation: “Describe the problem with your equipment.”
Host application: Sends the collected description to its chosen AI service, if appropriate.
Host application: Reviews the interpretation and applies its own business rules.
Andamento: Continues the configured interaction using application-supplied data or decisions.
BEYOND VOICE-ONLY INTERFACES
A native application can combine its own visual and touch interface with conversational interaction. For example, it may update a form or display results in response to information collected during a conversation.
Andamento coordinates the configured conversation; the host application owns screen rendering, touch input, and UI animation.
This is an application integration pattern, not a claim that Andamento automatically controls or animates UI elements or provides universal simultaneous-input arbitration.
ILLUSTRATIVE APPLICATION EXPERIENCE
THE BUSINESS CASE
See which conversation mechanics your team can reuse—and which application rules, interfaces, and integrations remain yours to build.
FOR DEVELOPMENT TEAMS