🏠 Overview › 🧩 Components ›🧩 Speech-to-Text Service

Speech-to-Text Service

Component Detail

Service Layer medium complexity mobilefrontend
0
Dependencies
0
Dependents
0
Entities
0
Integrations

Description

Wraps the platform speech recognition APIs and exposes a uniform dictation interface to the widget. Handles session lifecycle, language selection, and post-processing of recognised text before it is returned to the report composer.

Feature: Speech-to-Text Input

speech-to-text-service

Sources & reasoning

Priority matrix row 102 places Tale-til-tekst at SHOULD HAVE / Fase 2, which maps to the Core Product phase per §5. Blindeforbundet's anti-recording constraint is binding and drives the implementation boundary.

  • Tale-til-tekst i rapportskriving (Blindeforbundet + HLF) ... Blindeforbundet understreker at opptak under selve samtalen er uønsket - tale-til-tekst er for etterpå, ved rapportskriving.
  • Tale-til-tekst | - | - | ✓ | ✓ | ✓ | SHOULD HAVE | 2

Responsibilities

  • Initialise and tear down platform speech recognition sessions
  • Select Norwegian locale and apply punctuation/capitalisation post-processing
  • Emit interim and final transcript events to subscribers
  • Enforce the post-visit-only constraint by refusing to start during in-conversation flows
  • Report capability and permission status to callers

Interfaces

startSession(options): SessionHandle
stopSession(handle): FinalTranscript
subscribe(handle, listener): Unsubscribe
checkCapability(): CapabilityStatus