Toshius Klay

TypeScript, Python, Next.js, Whisper, Speech Recognition, STT, Audio AI, PostgreSQL, Node.js

Multi-speaker diarization under real-world noise. On-device STT inference optimisation. LLM post-processing for meeting intelligence.

Whisper-based on-device transcription + meeting intelligence layer at Wisprs. Multi-speaker diarization under real-world noise conditions.

Questions about STT, Whisper, audio pipelines, and meeting AI. Say hi if you're building voice-first tools or working on transcription accuracy.