How the SpeechTiles Documentation Fits Together
The SpeechTiles Andamento documentation is organized by the question you are trying to answer: what API do I call?, what configuration do I provide?, or how do I integrate the complete workflow?
SDK Reference — Swift interfaces
- STConversationRunner: lifecycle, resource loading, runtime fields, user intents, callbacks, and conversation control.
- STMediaRunner: direct playback, recording, speech synthesis, speech recognition, media parameters, events, and logging.
Method pages document Swift declarations, parameters, results, usage, and related APIs. For overloaded methods, display labels such as (URL) and (JSON) identify the variant; the Swift declarations remain authoritative.
Configuration Reference — declarative resources
- Conversation Configuration Reference: modules, transitions, recording and recognition behavior, and execution flow.
- Prompt Module Configuration Reference: reusable prompt and media definitions.
- Audio Resource to Text Mapping: supporting resource formats, including audio-to-text mappings.
These resources are represented in JSON, but the important distinction is their role in the running conversation—not the serialization format.
Guides and Samples — complete workflows
Use the Conversation quick start, Media quick start, Configuration and Runtime guide, and Samples to connect API calls, configuration, permissions, and callbacks.
Example: where to look for a prompt-loading question
To learn how to load prompt modules, read loadAudioPromptModules (URL). To learn which fields belong in those modules and how playback is configured, read the Prompt Module Configuration Reference. To understand how the conversation executes them, read Conversation Configuration Reference and runConversation.
Operational semantics
Consult STResult and the relevant API page to distinguish synchronous completion from asynchronous request acceptance. Use STConversationCallback and the Callback Semantics guide for callback timing and synchronous application updates.