xAI Senior Software Engineer, Media at SpaceXSI to build large-scale video services and real-time audio/video calling infrastructure. Focus on low-latency, resilient media delivery and integrating AI features into calls.
Responsibilities
Build the next generation of large-scale video services on X, spanning on-demand playback, live streaming, Spaces, and real-time audio/video calls
Design, build, and scale real-time audio/video calling infrastructure, including 1:1 and multi-party calls, using WebRTC and related real-time protocols (RTP/RTCP, SRTP, ICE/STUN/TURN)
Architect and operate media servers (SFU/MCU) that route, mix, and forward streams for many concurrent participants, and scale them globally to handle huge numbers of simultaneous calls
Deliver low-latency, resilient media under real-world network conditions through congestion control, adaptive bitrate, simulcast/SVC, jitter buffering, packet-loss recovery (FEC, NACK/RTX), and fast reconnection
Work hands-on with audio and video codecs (e.g. Opus, H.264, VP8/VP9, AV1) and hardware acceleration to balance quality, latency, bandwidth, and battery across devices
Build core calling features such as screen sharing, group calls, call recording, and seamless switching across iOS, Android, and web
Bring AI into the call experience: real-time transcription and captions, live translation, noise suppression and echo cancellation, video enhancement, and other intelligent in-call features powered by SpaceXSI models
Contribute to and rebuild core media processing pipelines in high-performance languages (Rust/C++)
Obsess over every millisecond and pixel, ensuring end-to-end media quality and performance at scale across a rich suite of products and user platforms; instrument call quality metrics (latency, jitter, packet loss, freeze rate, MOS) and use them to drive continuous improvement
Qualification
Experience designing
Required
5+ years of experience building high-performance media processing pipelines
Preferred
Proficient in high-performance Rust or C++
In-depth knowledge and experience building high-performance media processing pipelines
Hands-on experience building real-time audio/video communication products (e.g. video conferencing, VoIP, or live interactive streaming) at scale
Deep familiarity with WebRTC internals and real-time networking: RTP/RTCP, ICE/STUN/TURN, congestion control (e.g. GCC, transport-wide CC), and bandwidth estimation
Experience designing, deploying, or scaling SFU/MCU media servers for multi-party calls with many concurrent sessions
Strong understanding of audio/video codecs (Opus, H.264, VP8/VP9, AV1), encoding/decoding trade-offs, and hardware acceleration on mobile and desktop
Experience improving call quality and resilience under poor network conditions: packet loss, jitter, and bandwidth fluctuation
Experience with audio processing (echo cancellation, noise suppression, automatic gain control) or integrating ML models into real-time media, such as live transcription, translation, or video enhancement