Founding Voice AI Engineer (Part-time, potential path to Chief AI Officer)
Core
Build AI systems for pronunciation intelligence (Euphonia) to fix name and out-of-vocabulary word errors in voice agents, including fine-tuning G2P models, building validation pipelines, and creating context-aware ranking systems.
Role type
Founding Voice AI Engineer (Part-time, potential path to Chief AI Officer)
Builds
Drop-in pronunciation layer for voice AI agents, SDKs/MCP integrations, and internal AI tooling
Domain
Voice AI, Speech Recognition, Natural Language Processing
Deliverable
production ML models
Required skills
Fine-tuning G2P and pronunciation models, building multi-model AI validation pipelines, designing data architecture for verified pronunciation entries, developing context-aware ranking systems, AI-assisted development (Claude Code, Cursor, agent frameworks), hand-transcribing IPA
Preferred skills
Building agents and automations, defining pronunciation benchmarks, publishing technical papers
Technologies
G2P, phonetic transcription, LLMs, TTS providers, agent frameworks, RAG systems
Responsibilities
Fine-tune pronunciation models on proprietary dataset, build SDKs for voice infrastructure integrations, implement customer-specific fine-tuned models, define pronunciation reliability benchmarks, build internal AI tooling and agents
Seniority
Founding, hands-on IC with strategic influence
