NVIDIA AI Inference Engineer – GPU Data Center

Asiacruit BPO, Inc.; Jakarta Pusat; On-site; IDR 700,000,000 - 900,000,000; Full time

Job summary

Asiacruit BPO, Inc. in Jakarta is seeking an NVIDIA AI Software Engineer to design and optimize AI inference stacks across local and cloud-based GPU environments. You will work with PyTorch, vLLM, llama.cpp, and TensorRT, and you will integrate CUDA, NeMo, and NIM into robust workflows.

You will tackle topology-aware scheduling, GPU workload optimization, and runtime troubleshooting, documenting configurations and deployment strategies for efficient AI workloads.

Qualifications

• 5+ years of experience building and optimizing AI inference stacks in local, cloud, or multi-GPU environments.
• Hands-on experience with AI frameworks or inference technologies such as PyTorch, vLLM, llama.cpp, or TensorRT.
• Working knowledge of CUDA and the NVIDIA AI software ecosystem.
• Familiarity with NVIDIA NeMo, NIM, DGX, RTX, or related technologies.
• Understanding of AI inference deployment, GPU performance optimization, and workload execution.
• Experience with topology-aware scheduling tools is an advantage.
• Strong technical troubleshooting and problem-solving skills.
• Good command of English (mandatory).

Responsibilities

• Build and optimize local or cloud-based AI inference stacks for NVIDIA RTX, DGX, and multi-GPU environments.
• Work with modern AI frameworks and execution technologies such as PyTorch, vLLM, llama.cpp, and TensorRT.
• Support AI software and tooling, including CUDA, NVIDIA NeMo, and NVIDIA NIM.
• Contribute to topology-aware scheduling and GPU workload optimization.
• Investigate inference performance bottlenecks and improve GPU utilization, throughput, and latency.
• Troubleshoot software, framework, and runtime issues across AI environments.
• Document technical configurations, deployment practices, and optimization approaches.

Skills

AI inference stacks

PyTorch

vLLM

llama.cpp

TensorRT

CUDA

NVIDIA NeMo

NVIDIA NIM

Topology-aware scheduling

GPU performance optimization

Troubleshooting

English (mandatory)

Tools

CUDA toolkit

Job description

Asiacruit BPO, Inc. in Jakarta is seeking an NVIDIA AI Software Engineer to design and optimize AI inference stacks across local and cloud-based GPU environments. You will work with PyTorch, vLLM, llama.cpp, and TensorRT, and you will integrate CUDA, NeMo, and NIM into robust workflows.

You will tackle topology-aware scheduling, GPU workload optimization, and runtime troubleshooting, documenting configurations and deployment strategies for efficient AI workloads.