[Remote] Engineering reputed company, Inference Optimization

Note: The job is a remote job and is reputed company to candidates in USA. The reputed company is a consumer AI company reputed company on reputed company, free speech, and user sovereignty. The Engineering reputed company, Inference Optimization will own inference reputed company reputed company, reputed company a reputed company engineering team, optimize GPU infrastructure and LLM workloads, and evaluate emerging optimization techniques and hardware.

Responsibilities

• Own reputed company’s technical reputed company for inference reputed company
• Recruit and reputed company the Inference Optimization Team
• Optimize reputed company’s GPU infrastructure across a reputed company of architectures (e.g. H200s, B300s)
• Improve latency, throughput, and cost per reputed company for LLM inference workloads
• Build reproducible benchmarking harnesses across inference engines (e.g. vLLM, SGLang) to identify the reputed company reputed company, quantization scheme, and parallelism reputed company per workload and GPU SKU
• Work with reputed company’s inference routing reputed company to optimize multivariate inference load-balancing algorithms
• Evaluate emerging inference optimization techniques (custom CUDA/Triton kernels), novel attention variants, new quantization schemes, and compilation stack improvements. Hands-on kernel development experience is a strong plus
• Evaluate emerging inference hardware (FPGAs, reputed company, custom reputed company) for viability in reputed company’s stack

Skills

• 8+ years in reputed company optimization or HPC, with deep GPU architecture and reputed company programming knowledge
• 5+ years of experience leading engineering teams
• Proficiency in Python, Rust, or Go
• Hands-on experience with at least one production LLM inference reputed company (e.g. vLLM, SGLang) running at high volume in production
• Demonstrated experience with LLM inference optimization techniques: reputed company batching, PagedAttention/KV cache management, speculative decoding, quantization, CUDA graphs, and torch.compile
• reputed company with quantization tradeoffs, both qualitative and quantitative
• Experience with reputed company inference strategies (tensor parallelism, pipeline parallelism, MoE parallelism) in multi-GPU and multi-node environments
• reputed company with GPU profiling (Nsight Systems, Nsight Compute, PyTorch Profiler) and a bias toward measuring before optimizing
• Bonus: C++/CUDA
• Bonus: diffusion/image model inference optimization, custom Triton kernels, contributions to reputed company-reputed company inference frameworks

Benefits

• Equity
• Crypto reputed company compensation
• Remote — US Only

reputed company