Solution Architect

• About Lablup

  We are a company providing machine learning platforms and solutions under the motto 'Make AI Accessible'. Lablup strives to solve various problems that arise when sharing scientific and engineering research processes through a standardized platform, and automates all stages of research and service. We develop and service Backend.AI to realize this vision. Backend.AI, a hyperscale AI platform, reduces the difficulties in developing and serving various AI models, from small-scale AI models to large language models.

• About Solution Architect at Lablup

  Lablup's Solution Architects are key technical experts who design AI infrastructure strategies for customers and guide them with technical direction. Engaging from the initial evaluation stage, they analyze customers' requirements and existing environments and propose optimal solution architectures that account for performance, scalability, security, and timeline together.

  If Solution Engineers make the platform run reliably within the customer's environment, Solution Architects work one step upstream, defining what to build and how to build it. They must also be able to stand up PoC environments and run benchmarks themselves, personally validating on real hardware that their proposed architecture holds up.

  This is a role where you design some of the largest financial and public-sector GPU cluster deployments at home and abroad from the very front line, and feed the insights gained in the field directly into the product direction.

• Key Responsibilities
  • Requirements analysis and solution design: Define the optimal Backend.AI configuration and technology stack from the customer's AI goals and existing infrastructure.
  • Upfront capacity planning: Size GPU count, interconnect bandwidth, storage throughput/IOPS, and power/thermal/rack layout from workload characteristics, with clear rationale.
  • Technical proposals and consulting: Frame the performance, scalability, security, and schedule trade-offs for decision-makers, and advise on AI adoption and operations strategy.
  • RFP and technical specification response: Review RFPs and technical specs, author the technical sections of proposals, and handle technical review meetings.
  • Leading PoCs and deployments: Design and run PoCs, then work with engineering during build to carry the design intent into implementation.
  • Customer relationship management: Track key accounts, spot the right timing for expansion and scale-out, and align the technical roadmap with customers.
  • Monitoring and optimization: Diagnose performance bottlenecks across the platform and its dependent stack, and drive improvements.
  • Contribution to product improvement: Turn field requirements and market trends into input for the product roadmap.
  • Research and sharing of technology trends: Research emerging tech and best practices, and share them with customers and internal teams.
• Basic Qualifications
  • 7+ years of hands-on experience in IT infrastructure or platforms (in-house system operation alone doesn't qualify).
  • A track record of proposing technical solutions to customers or other teams and driving them to consensus.
  • Deep understanding of Linux networking, virtualization, and container technologies.
  • Experience designing or operating container orchestration platforms (e.g., Kubernetes).
  • Understanding of distributed training/inference: multi-node communication (e.g., NCCL), resource isolation and multi-tenancy, and scheduling.
  • Ability to structure problems and propose alternatives when requirements are undefined.
  • Able to travel frequently to domestic and overseas customer sites.
  • Able to read and write English technical documentation comfortably.
• Preferred Qualifications
  • LLM inference serving stacks (vLLM, SGLang, TensorRT-LLM): sizing requirements by model size, quantization, and concurrency.
  • GPU cluster design with high-speed fabrics (InfiniBand, RoCEv2) and parallel/HPC file systems (WEKA, DDN, VAST, Lustre, GPFS).
  • High-availability builds for air-gapped or tightly secured customers (public, finance, defense), working with distributors, SIs, and hardware vendors, including network separation and security reviews.
  • Hands-on use and operation of AI/ML platforms (Kubernetes, Docker, PyTorch).
  • Enterprise (B2B) pre-sales technical consulting.
  • CI/CD pipelines with DevOps and automation tools (Ansible, Terraform, Pyinfra).
  • Able to hold technical discussions in spoken English.
• You'll be a great fit if you
  • Can structure open-ended requirements and propose a specific configuration with clear reasoning.
  • Take ownership through to seeing your design validated on real hardware.
  • Keep emotion and fact separate under customer pressure, and speak the right language for both decision-makers and on-site engineers.
  • Openly admit what you don't know and work to find a path forward.
• This role may not be for you if you
  • Prefer implementing predefined requirements.
  • Find customer-facing work, proposals, and negotiation burdensome.
  • Expect a primarily remote/work-from-home role.
• Benefits and Perks
  • Fitness stipend
  • Welfare points
  • Lunch meal support
  • Complimentary snacks always available
  • Annual comprehensive or semi-comprehensive health checkup for employee and spouse
  • Opportunities to participate in developer communities and industry events
  • Employee benefits as a certified family-friendly workplace by the Ministry of Gender Equality and Family