- Take architectural ownership of open-source libraries like speech-to-speech, focusing on pipeline design and real-time latency budgets.
- Integrate new ASR, TTS, and end-to-end speech models while maintaining clean abstractions.
- Design and implement developer APIs, streaming protocols, and session lifecycle management for hf-voice.
- Build and optimize real-time GPU inference, including concurrency, autoscaling, and observability.
- Take products from demo stage to production, establishing SLOs and load testing protocols.
- Engage with the community by reviewing PRs, triaging issues, and creating documentation, examples, and templates.