bodhi-realtime-agent / STTProvider

Interface: STTProvider

Defined in: src/types/transport.ts:146

Provider-agnostic interface for pluggable speech-to-text providers.

VoiceSession creates the provider, calls configure() with the transport's audio format, then start(). Audio flows via feedAudio(); turn signals via commit()/handleInterrupted()/handleTurnComplete(). Results arrive via the onTranscript/onPartialTranscript callbacks.

Properties

onPartialTranscript?

optional onPartialTranscript?: (text) => void

Defined in: src/types/transport.ts:190

Partial/interim transcription (streaming providers only). Replaces any previous partial for the same turn.

Parameters

text

string

Returns

void

---

onTranscript?

optional onTranscript?: (text, turnId) => void

Defined in: src/types/transport.ts:186

Final transcription of user speech.

Parameters

text

string

The transcribed text.

turnId

number | undefined

The turn this transcript belongs to (from commit()). Undefined when a streaming provider's VAD auto-commits before the framework calls commit().

Returns

void

---

supportedEncodings?

readonly optional supportedEncodings?: readonly ("pcm" | "pcmu")[]

Defined in: src/types/transport.ts:151

Optional static declaration of which audio encodings the provider can consume. When omitted, defaults to ['pcm'] (today's behaviour). VoiceSession reads this at configure time and decides whether to feed raw PCM or encode to G.711 before calling feedAudio().

Methods

commit()

commit(turnId): void

Defined in: src/types/transport.ts:171

Signal that the user's turn has ended (model started responding). For batch providers, this triggers transcription. For streaming providers, this may trigger a manual commit.

Parameters

turnId

number

Monotonically increasing turn counter for ordering.

Returns

void

---

configure()

configure(audio): void

Defined in: src/types/transport.ts:156

Configure the audio format that feedAudio() will deliver. Called once before start(). The provider MUST resample or reject if it cannot handle the given format.

Parameters

audio

STTAudioConfig

Returns

void

---

feedAudio()

feedAudio(base64Pcm): void

Defined in: src/types/transport.ts:165

Feed audio data. Format matches the STTAudioConfig from configure().

Parameters

base64Pcm

string

Base64-encoded PCM audio chunk.

Returns

void

---

handleInterrupted()

handleInterrupted(): void

Defined in: src/types/transport.ts:175

Signal that the current turn was interrupted by the user. Providers MUST preserve buffered audio for the next commit().

Returns

void

---

handleTurnComplete()

handleTurnComplete(): void

Defined in: src/types/transport.ts:179

Signal a natural turn completion (model finished, no interruption). Batch providers SHOULD clear buffers. Streaming providers may no-op.

Returns

void

---

start()

start(): Promise<void>

Defined in: src/types/transport.ts:159

Start the STT session (e.g. open WebSocket).

Returns

Promise<void>

---

stop()

stop(): Promise<void>

Defined in: src/types/transport.ts:161

Stop the STT session (e.g. close WebSocket).

Returns

Promise<void>