Senior Software Engineer

What you'll own

• Ship features end-to-end: take loosely defined product priorities (e.g. voice localization for accent support, multilingual sentence endpointing) and turn them into specs, architecture, and shipped features.
• Partner with Research to bring state-of-the-art real-time models into the Conversational Video Interface (CVI). This may involve dipping into research territory.
• Fix issues proactively: address flaky processes or slow paths without needing tickets or permission.

What this role has shipped

• Guided Browser Control: framework allowing PALs to drive a real web browser mid-conversation, narrating clicks and scrolls.
• Magic Canvas: interactive cards rendered on the video stream (questions, text boxes, calendars, charts, alerts).
• Memories and Knowledge Base: context layer for recalling past conversations and pulling from document stores in real time.
• Utterance-to-utterance optimization: reducing conversational latency by hundreds of milliseconds.

Requirements

• Fluent in Python with deep expertise in concurrent programs (asyncio, multiprocessing, event loops, GIL, shared memory).
• Experience building real-time systems (video, audio, or model inference) and working within millisecond latency budgets.
• Senior track record of setting direction, owning decisions, and delivering results.
• Ability to explain complex concepts clearly and teach others.

Nice-to-haves

• Experience with GPUs and GPU management.
• Deep experience with video streaming over WebRTC.
• Deep experience with LLMs and agentic frameworks.
• Extensive experience with low-level systems.