L
AI Engineer
LiveX AI, Inc.
5 hours ago
Full-time
On-site
Palo Alto, California, United States
LiveX AI is developing real-time, interactive AI avatars and is hiring an AI/Machine Learning Engineer to train, fine-tune, and serve generative models end to end. The position emphasizes low-latency inference for live streaming and supports the surrounding data, training, evaluation, and deployment infrastructure.
Responsibilities
Train and fine-tune generative models for talking-head synthesis, audio-driven facial animation, and full-body avatar video, including diffusion models, DiT, 3D Gaussian Splatting, GAN-based approaches, and transformer-based video generators.
Design data pipelines for large-scale video, audio, and multimodal datasets, covering collection, cleaning, labeling, alignment, and augmentation.
Optimize models for real-time serving with sub-200ms end-to-end latency using techniques such as distillation, quantization, pruning, caching, streaming inference, and custom CUDA/Triton kernels when needed.
Build and maintain training, evaluation, and deployment infrastructure on GPU clusters using PyTorch, DeepSpeed, FSDP, Ray, and Kubernetes.
Define and track quality metrics, including lip-sync accuracy, identity preservation, expression naturalness, latency, and throughput, and drive measurable improvements.
Partner with backend, product, and research teams to ship avatar features and iterate based on user feedback.
Stay current with research in generative video, multimodal AI, and neural rendering, integrating useful techniques into production systems.
Requirements
3+ years of hands-on ML engineering, or strong research experience with production-quality code, with a history of shipping deep learning models.
Deep expertise in PyTorch and modern deep learning workflows, including the ability to read, reproduce, and extend research papers.
Strong background in one or more areas such as diffusion models, video generation, generative AI, computer vision, multimodal (audio plus vision) AI, speech synthesis, or neural rendering.
Solid Python engineering, including distributed training across multi-GPU or multi-node setups.
Proven real-time inference optimization experience using TensorRT, ONNX, Triton Inference Server, CUDA, mixed precision, and streaming pipelines.
Comfort working in ambiguous, fast-changing environments with proactive, ownership-driven execution.
Startup mindset focused on pushing the capabilities of interactive avatars.
Technologies
PyTorch
DeepSpeed
FSDP
Ray
Kubernetes
CUDA
Triton
Triton Inference Server
TensorRT
ONNX
Python
Benefits
Competitive salary, equity, and benefits
Access to substantial GPU compute
Nice-to-Haves
Experience related to talking-head systems, audio-driven animation, lip-sync, or digital human systems, including work with 3D Gaussian Splatting, NeRF, face reenactment, or avatar synthesis.
Publications at top venues (NeurIPS, CVPR, ICCV, SIGGRAPH, ICLR, ICML) or notable open-source contributions.
Compensation Compensation includes
competitive salary, equity, and benefits, plus access to substantial GPU compute. No range is listed. LiveX AI is an Equal Opportunity Employer.
Location and Experience
Location:
Palo Alto, CA (onsite)
Minimum experience:
3 years
#J-18808-Ljbffr
Responsibilities
Train and fine-tune generative models for talking-head synthesis, audio-driven facial animation, and full-body avatar video, including diffusion models, DiT, 3D Gaussian Splatting, GAN-based approaches, and transformer-based video generators.
Design data pipelines for large-scale video, audio, and multimodal datasets, covering collection, cleaning, labeling, alignment, and augmentation.
Optimize models for real-time serving with sub-200ms end-to-end latency using techniques such as distillation, quantization, pruning, caching, streaming inference, and custom CUDA/Triton kernels when needed.
Build and maintain training, evaluation, and deployment infrastructure on GPU clusters using PyTorch, DeepSpeed, FSDP, Ray, and Kubernetes.
Define and track quality metrics, including lip-sync accuracy, identity preservation, expression naturalness, latency, and throughput, and drive measurable improvements.
Partner with backend, product, and research teams to ship avatar features and iterate based on user feedback.
Stay current with research in generative video, multimodal AI, and neural rendering, integrating useful techniques into production systems.
Requirements
3+ years of hands-on ML engineering, or strong research experience with production-quality code, with a history of shipping deep learning models.
Deep expertise in PyTorch and modern deep learning workflows, including the ability to read, reproduce, and extend research papers.
Strong background in one or more areas such as diffusion models, video generation, generative AI, computer vision, multimodal (audio plus vision) AI, speech synthesis, or neural rendering.
Solid Python engineering, including distributed training across multi-GPU or multi-node setups.
Proven real-time inference optimization experience using TensorRT, ONNX, Triton Inference Server, CUDA, mixed precision, and streaming pipelines.
Comfort working in ambiguous, fast-changing environments with proactive, ownership-driven execution.
Startup mindset focused on pushing the capabilities of interactive avatars.
Technologies
PyTorch
DeepSpeed
FSDP
Ray
Kubernetes
CUDA
Triton
Triton Inference Server
TensorRT
ONNX
Python
Benefits
Competitive salary, equity, and benefits
Access to substantial GPU compute
Nice-to-Haves
Experience related to talking-head systems, audio-driven animation, lip-sync, or digital human systems, including work with 3D Gaussian Splatting, NeRF, face reenactment, or avatar synthesis.
Publications at top venues (NeurIPS, CVPR, ICCV, SIGGRAPH, ICLR, ICML) or notable open-source contributions.
Compensation Compensation includes
competitive salary, equity, and benefits, plus access to substantial GPU compute. No range is listed. LiveX AI is an Equal Opportunity Employer.
Location and Experience
Location:
Palo Alto, CA (onsite)
Minimum experience:
3 years
#J-18808-Ljbffr