Saylion: Heartland AI Bot

S

Saylion

Voicebot Assistant

Architectures

Cascade

ASR → LLM → TTS

SpeechLLM

End-to-end audio → text

            Both run in parallel on every turn.
        

Voice Input

TTS Engine

TTS Playback

            Voice responses enabled
        

Voice    Kokoro / Google
ASR      Speechlab
LLM      SEA-LION / DiVA

User Input

Voice

Waiting for input…

Cascade

ASR→LLM→TTS

Response will appear here…

Total —

ASR —

LLM —

TTS —

Play —

SpeechLLM

DiVA

Response will appear here…

Total —

Model —

TTS —

Play —

Click the mic to start recording.

                Cascade: text → LLM → TTS  |  SpeechLLM: text → TTS audio → model
            