GPU/ML Systems Engineer

Responsibilities

• Deploy and tune vLLM with multi-GPU tensor parallelism, dynamic batching, and KV cache optimization for LLMs
• Configure NVIDIA Triton for production multi-model serving with custom backends and model ensembles
• Build TensorRT-LLM optimized model binaries for maximum throughput on L40S, A100, and H100 GPUs
• Implement AWS Inferentia deployments using Neuron SDK, including model compilation and performance tuning
• Execute model quantization (INT8, FP16, GPTQ, AWQ) with rigorous quality-accuracy tradeoff analysis
• Run comprehensive load testing using Locust to map performance cliffs and scaling thresholds
• Produce detailed benchmark reports with instance selection and cost-per-token recommendations

Requirements

• 3–7 years of experience with GPU-accelerated ML workloads in production
• Hands-on experience with LLM serving frameworks such as vLLM, TensorRT-LLM, or Triton Inference Server
• Deep understanding of GPU architecture, including memory hierarchy, tensor cores, NVLink, and NCCL
• Proficiency in model quantization techniques (INT8, FP16, GPTQ, AWQ)
• Strong knowledge of the CUDA ecosystem (drivers, cuDNN, NVIDIA container toolkit)
• Experience with performance profiling tools like Nsight, nvidia-smi, or DCGM
• Practical experience managing AWS GPU instances (G-series, P-series)

Preferred Qualifications

• Experience optimizing models for custom accelerators like AWS Inferentia or Trainium
• Familiarity with KServe and Prometheus + DCGM Exporter for monitoring

Benefits

• Learn from experts, including former AWS leaders and AI pioneers
• Direct ownership of high-impact "greenfield" projects from concept to launch
• Access to modern tech stacks, including the latest Generative AI frameworks
• Opportunity for rapid career growth in a high-speed environment

About the Company

Aivar Innovations is an AI-first technology partner where cutting-edge technology meets industry expertise. We provide AI-augmented teams that accelerate development, reduce time-to-market, and deliver exceptional code quality for major global enterprises.