CAREERS / TOKYO
Build the foundation for AI.
From networking to model optimization. Join the engineering team behind Token Factory in Tokyo.
Explore open rolesTokyo · On-site
Nexora is an AI infrastructure company focused on deploying AI model inference environments and providing a compute rental platform. More than 50 colleagues across Japan, the United States and China work in business, R&D and operations. We are hiring AI technology and infrastructure specialists in Japan.
Colleagues across Japan, the US & China
50+
WORKING AT NEXORA
Conditions for all roles
- Location
- Tokyo (on-site)
- Annual salary
- JPY 10 million or more, to be discussed during interviews
- Employment
- Permanent or contract employment
- Probation
- 3 months
- Languages
- Chinese, plus professional communication skills in either English or Japanese
- Experience
- No minimum number of years. We value technical ability, project experience and execution.
OPEN POSITIONS
Open roles
Multiple openings in each role
NETWORKNetwork Engineer
Responsibilities
- Own the architecture, implementation planning and deployment of Ethernet and external connectivity networks for GPU clusters in data centers.
- Plan network capacity, redundancy and expansion based on business requirements.
- Operate and monitor networks, improve performance, and investigate and respond to incidents.
- Work with R&D and operations teams to support infrastructure deployment and service launches.
Requirements
- Ability to independently design data center network architecture and drive implementation.
- Practical experience building, operating and troubleshooting data center Ethernet and external connectivity networks.
- Knowledge and experience configuring and maintaining data center switches. We primarily use H3C and Ruijie products.
- Ability to balance performance, reliability, scalability and cost against business requirements.
Preferred qualifications
- Practical experience with InfiniBand, especially building and operating networks for GPU clusters.
SERVER OPERATIONSServer Operations Engineer
Responsibilities
- Handle server rack installation, hardware configuration, OS installation, handover acceptance and routine maintenance.
- Monitor servers, investigate incidents and restore service; identify and resolve hardware, OS and runtime problems.
- Manage server assets, configurations and maintenance records, and maintain operations documentation.
- Standardize and automate server deployment, inspection and incident response.
- Coordinate with networking, R&D and data center teams to operate the infrastructure behind AI inference environments and the compute rental platform.
Requirements
- Strong Linux administration skills, including server deployment, maintenance and troubleshooting.
- Knowledge of core server hardware and ability to investigate common CPU, memory, disk and NIC faults.
- Basic networking knowledge and ability to assist with server connectivity and access issues.
- Ability to write clear incident records and operational documentation and collaborate with a team.
Preferred qualifications
- Experience operating GPU servers, installing GPU drivers and troubleshooting related issues.
- Experience with bulk server deployment, operations automation, or monitoring and alerting systems.
- On-site data center operations and maintenance experience.
DATA CENTER DELIVERYData Center Construction & Delivery Manager
Responsibilities
- Manage delivery of new and expanded data centers, from requirements and preparation through construction coordination, equipment delivery, installation, commissioning, acceptance and operational handover.
- Build project plans defining scope, milestones, resources and acceptance criteria; manage progress, cost and quality.
- Coordinate data center operators, design firms, contractors and equipment vendors to deliver power, cooling, racks and cabling to plan.
- Work with internal networking, server and R&D teams to align facility construction and IT equipment deployment for service launch.
- Manage risks, dependencies and requirement changes, resolve on-site issues and report project status.
- Coordinate testing and acceptance, track corrective actions, and hand over as-built documentation, asset lists and operational documents.
Requirements
- Practical experience with new data centers, refurbishment or expansion, with the ability to drive projects from planning through acceptance and handover.
- Understanding of power, cooling, racks, structured cabling and network connectivity, including system interfaces and responsibility boundaries.
- Ability to manage schedule, cost, quality and risk and coordinate multiple vendors and internal teams.
- Ability to review implementation plans, track on-site progress, coordinate acceptance and resolve issues.
- Clear documentation, project reporting and stakeholder coordination skills.
Preferred qualifications
- Experience delivering and handing over data center projects in Japan.
- Project experience involving GPU clusters, high-density racks or liquid cooling.
- Experience managing data center operators, contractors and equipment vendors.
- PMP or other project management or data center qualifications.
MODEL OPTIMIZATIONModel Optimization Engineer
Responsibilities
- Deploy and optimize AI inference to improve throughput, reduce latency and GPU memory use, and improve GPU resource utilization.
- Select and tune inference frameworks for each use case; work on quantization, operator optimization, batching and caching strategies.
- Build performance and model quality evaluation processes, varying hardware, model size and concurrency to balance performance, cost and output quality.
- Identify compute, GPU memory, communication and scheduling bottlenecks and implement improvements.
- Work with platform development, server and network teams to integrate optimization results into inference services and the compute platform.
- Document deployment, testing and optimization, and build reusable tools and processes.
Requirements
- Understanding of deep learning fundamentals, common architectures such as Transformers, and inference execution.
- Proficiency in Python and frameworks such as PyTorch, with Linux development and troubleshooting skills.
- Practical experience building and optimizing inference environments, independently analyzing bottlenecks, implementing improvements and measuring results.
- Experience with an inference framework or tool such as vLLM, TensorRT, TensorRT-LLM or ONNX Runtime.
- Understanding of GPU computing and memory management, and the relationships between numerical precision, memory use, throughput and latency.
Preferred qualifications
- Experience optimizing LLM or multimodal model inference.
- Practical experience with low-precision quantization, pruning or knowledge distillation.
- Knowledge of C++, CUDA or Triton and experience optimizing custom operators or GPU kernels.
- Experience with multi-GPU or multi-node inference, or distributed communication optimization.
- Contributions to open-source inference frameworks, model optimization tools or related projects.
JOIN NEXORA
How to apply
Send your résumé and work history to the email address below. Include the position you are applying for in the subject line.