BrainDance

Devlog

• Sprint 2 (Weeks 3–4): Determinism, STT Control, and State Introduction
  May 12, 2026 by Kevsta4Thumbs

  This sprint was dedicated to restoring determinism. On the TTS side, WAV decoding was validated and voice selection restricted to known‑good IDs to prevent zero‑length audio. On the STT side, long... Continue reading

  
• Sprint 1 (Weeks 1–2): Speech Pipeline Proof‑of‑Concept and Concurrency Failure
  May 12, 2026 by Kevsta4Thumbs

  The initial sprint focused on validating whether a full real‑time speech pipeline could operate inside Unity. Whisper STT, a local LLM, and TTS were integrated end‑to‑end. From a purely function... Continue reading

  
• Week 1- Tech Discussion
  May 12, 2026 by Kevsta4Thumbs

  Core Speech Pipeline Integration The initial technical focus was establishing a complete speech pipeline inside Unity. I integrated Whisper for speech‑to‑text, a local language model for dialogue... Continue reading