Deloitte

Audits companies and advises on tax and technology

AI Supercompute Infrastructure Solutions Architect

Implementation & Technical Account ManagementConfirmed live in the last 24 hours

Full-timePolandNo salary listed

About the Job

Requirements

At least 4–8 years of hands-on experience in infrastructure, systems, platform, networking, cloud, high-performance computing, storage, automation, or data-centre engineering.
Strong practical, hands-on depth in at least one core infrastructure discipline, such as GPU and server infrastructure, enterprise or high-performance networking, storage, Linux and cluster platforms, cloud infrastructure, automation, or data-centre facilities.
A strong interest in developing breadth across compute, storage, high-performance networking, physical data-centre infrastructure, and cluster platforms; expert-level knowledge across every component is not expected.
Ability to work closely with specialists in adjacent domains, ask effective questions, understand cross-stack dependencies, challenge assumptions constructively, and turn multiple technical inputs into a coherent end-to-end solution.
Experience designing, deploying, operating, or supporting enterprise-grade server, storage, networking, private-cloud, high-performance computing, or data-centre environments.
Strong understanding of server architecture, including CPU, GPU, memory, PCI Express, Non-Volatile Memory Express, network interface cards, BIOS and firmware configuration, remote management, hardware lifecycle management, and fault isolation.
Good understanding of data-centre fundamentals, including rack design, power distribution, redundancy, power density, cooling, structured cabling, physical installation, and operational readiness.
Solid knowledge of networking fundamentals, including switching and routing, virtual local area networks, maximum transmission unit, link aggregation, IP addressing, traffic flows, latency, bandwidth, resiliency, and fault isolation.
Working knowledge of high-performance networking concepts and technologies, including InfiniBand, Remote Direct Memory Access, RDMA over Converged Ethernet, GPUDirect RDMA, or equivalent high-performance computing and low-latency interconnects.
Understanding of storage concepts relevant to AI and high-performance workloads, including block, file, object, and parallel file systems; NVMe-based storage; throughput; input/output operations per second; metadata; replication; backup; and data lifecycle.
Familiarity with Linux administration and troubleshooting, including system services, networking, storage, drivers, package management, logs, performance diagnostics, and automation.
Practical familiarity with Kubernetes, Slurm, or equivalent cluster platforms, including node onboarding, GPU scheduling, workload placement, networking, persistent storage, and day-two operational requirements.
Experience working with technology vendors, delivery partners, customers, or cross-functional engineering teams.
Ability to turn incomplete or ambiguous requirements into explicit technical decisions, architecture principles, design artefacts, implementation plans, and measurable acceptance criteria.
Strong written and verbal communication skills, including the ability to explain technical trade-offs to engineering teams, vendors, customers, and senior decision-makers.
Comfort working in data-centre environments and reviewing physical infrastructure, server hardware, racks, cabling, power, cooling, and installation quality.
Experience with infrastructure automation or Infrastructure as Code, such as Ansible, Terraform, Python, Bash, GitOps, or equivalent tooling.
Fluent English and Polish is required.

Responsibilities

Translate customer requirements, including AI training, fine-tuning, inference, retrieval-augmented generation, simulation, data volumes, performance targets, availability objectives, and future growth, into complete architectures across compute, storage, networking, data-centre infrastructure, and platform software.
Lead technical discovery, requirements gathering, architecture workshops, solution reviews, and design decisions. Establish the target-state architecture, identify dependencies and risks, define technical standards, and provide a path from concept through implementation.
Develop detailed, technically validated, and commercially viable bills of materials covering GPU servers, CPU and management nodes, storage systems, InfiniBand or Ethernet fabrics, switches, network interface cards, data processing units, optics, cables, racks, power distribution units, power infrastructure, cooling equipment, and relevant software, subscriptions, and support services.
Work directly with server original equipment manufacturers, GPU vendors, storage providers, networking vendors, system integrators, and specialist partners to validate configurations, confirm compatibility, align performance and resiliency assumptions, review lead times, compare options, and resolve technical dependencies.
Define infrastructure for AI training, inference, fine-tuning, large language model inference, and data-processing workloads. Size GPU capacity and memory, CPU, RAM, topology, local NVMe, host networking, management infrastructure, and expansion paths based on workload requirements.
Define high-performance storage architectures for AI and GPU-intensive workloads, covering capacity, throughput, input/output operations per second, metadata performance, caching, tiering, data protection, backup, replication, and integration with Kubernetes, Slurm, and other cluster platforms.
Define, with network specialists and technology vendors, high-performance InfiniBand and Ethernet/RDMA over Converged Ethernet fabrics for GPU clusters. Validate topology, non-blocking or oversubscribed design targets, switch placement, host connectivity, optics, cabling, routing, quality-of-service requirements, congestion management, and operational management needs.
Act as the technical liaison between customers, colocation providers, data-centre operators, facilities teams, installation contractors, and technology partners. Define and validate rack layouts, floor-space requirements, structural constraints, power density, A/B feeds, power distribution unit design, cooling capacity, airflow, liquid-cooling requirements, cable pathways, and commissioning plans.
Own the correctness of the delivered design. Define installation standards, review rack elevations and port maps, validate cabling plans, oversee data-centre teams and third-party contractors, and ensure physical implementation matches the approved architecture.
Partner with platform engineering teams to ensure physical infrastructure is ready for Kubernetes, Slurm, GPU scheduling, workload orchestration, multi-tenancy, identity and access integration, observability, security controls, node lifecycle management, and operational automation.
Define acceptance criteria and coordinate validation of the complete environment, including hardware health, GPU configuration, network connectivity and fabric health, storage performance, node configuration, benchmark execution, burn-in testing, operational documentation, and handover.
Create architecture documents, high-level and low-level designs, bills of materials, rack elevations, network diagrams, deployment plans, technical proposal content, and statements of work. Lead technical workshops, support requests for proposals and customer presentations, and provide technical leadership throughout the solution lifecycle.

Desired Qualifications

Experience with NVIDIA AI infrastructure, including DGX, HGX, NVIDIA RTX PRO Servers, BasePOD, SuperPOD, DGX SuperPOD, NVIDIA AI Enterprise, or NVIDIA reference architectures.
Exposure to NVIDIA networking products, including Mellanox and ConnectX network interface cards, BlueField data processing units, Quantum InfiniBand switches, Spectrum Ethernet switches, LinkX cabling, NVIDIA UFM, or NVIDIA NEO.
Familiarity with AI platform components such as Kubernetes, Slurm, NVIDIA GPU Operator, DCGM, NVIDIA Base Command Manager, Run:ai, Ray, Kubeflow, MLflow, vLLM, Triton Inference Server, or similar technologies.
Experience designing or sizing high-performance storage environments, including parallel file systems, NVMe over Fabrics, Lustre, DDN, VAST Data, NetApp, Pure Storage, or equivalent platforms.
Familiarity with GPU workload characteristics, distributed training, collective communication patterns, NCCL, inference serving, large language model fine-tuning, retrieval-augmented generation pipelines, and data-pipeline bottlenecks.
Experience with liquid-cooled infrastructure, direct-to-chip cooling, high-density GPU racks, or data-centre design for AI and high-performance computing workloads.
Hands-on experience with infrastructure validation or performance tools such as DCGM, NCCL tests, iperf3, perftest, ibdiagnet, fio, MLPerf, or vendor-specific diagnostic tooling.
Experience in consulting, technical presales, solution architecture, request-for-proposal delivery, or another customer-facing technical role.

Who we are looking for

We are a technology consulting firm building and operating next-generation AI supercompute infrastructure for the world’s most ambitious organizations. As an AI Infrastructure Solutions Architect, you will define and guide the delivery of complete AI supercomputing solutions.

You will work with clients to understand their AI workloads, performance objectives, growth plans, operational requirements, and facility constraints. You will translate these requirements into validated, implementation-ready architectures and detailed bills of materials across GPU compute, high-performance storage, networking, racks, power, cooling, and the platform stack required to operate clusters reliably at scale.

As a repeatedly awarded NVIDIA Consulting Partner of the Year in EMEA, we hold one of the deepest and most recognized NVIDIA partnerships in the region. This gives our architects privileged access to adoption programmes, reference architectures, and NVIDIA engineering teams.

You will work with technology—and at a scale—that most engineers will not encounter for years.

This is a role for an experienced platform, systems engineer who wants to move beyond executing predefined projects and take ownership of defining complete AI infrastructure solutions. You will shape technical decisions early, establish the design direction, align specialist vendors and delivery teams, and remain accountable for the solution through deployment and acceptance.

You have solid technical foundations and real hands-on experience. You understand that an AI supercomputer is not simply a collection of GPU servers: performance, resilience, and operational success depend on the interaction between compute architecture, storage throughput, network fabric, facility design, power density, cooling, and platform operations.

You do not need to arrive as an expert in every part of this stack. Very few people are. We are looking for someone with genuine practical depth in at least one relevant domain such as GPU and server infrastructure, high-performance networking, storage, Linux and cluster platforms or data-centre design and the curiosity, discipline, and ambition to build working depth across the remaining layers.

This role is for an engineer who wants to develop into an end-to-end AI infrastructure architect. You will work alongside domain specialists, leading technology vendors, data-centre teams, and experienced delivery engineers. In return, we expect intellectual curiosity, ownership, sound engineering judgement, the ability to ask the right questions, and a genuine desire to understand how every layer of an AI supercomputing environment works together.

Requirements:

At least 4–8 years of hands-on experience in infrastructure, systems, platform, networking, cloud, HPC, storage, automation, or data-center engineering.
Strong practical, hands-on depth in at least one core infrastructure discipline, such as GPU and server infrastructure, enterprise or high-performance networking, storage, Linux and cluster platforms, cloud infrastructure, automation, or data-centre facilities.
You are not expected to have expert-level knowledge across every component of AI supercomputing infrastructure. However, you should have a strong interest in developing breadth across compute, storage, high-performance networking, physical data-centre infrastructure, and cluster platforms.
Ability to work closely with specialists in adjacent domains, ask the right questions, understand cross-stack dependencies, challenge assumptions constructively, and turn multiple technical inputs into a coherent end-to-end solution.
Experience designing, deploying, operating, or supporting enterprise-grade server, storage, networking, private-cloud, HPC, or data-centre environments.
Strong understanding of server architecture, including CPU, GPU, memory, PCIe, NVMe, NICs, BIOS and firmware configuration, remote management, hardware lifecycle management, and fault isolation.
Good understanding of data-centre fundamentals, including rack design, power distribution, redundancy, power density, cooling, structured cabling, physical installation, and operational readiness.
Solid knowledge of networking fundamentals, including switching and routing, VLANs, MTU, link aggregation, IP addressing, traffic flows, latency, bandwidth, resiliency, and fault isolation.
Working knowledge of high-performance networking concepts and technologies, including InfiniBand, RDMA, RoCE, GPUDirect RDMA, or equivalent HPC and low-latency interconnects.
Understanding of storage concepts relevant to AI and high-performance workloads, including block, file, object, parallel file systems, NVMe-based storage, throughput, IOPS, metadata, replication, backup, and data lifecycle.
Familiarity with Linux administration and troubleshooting, including system services, networking, storage, drivers, package management, logs, performance diagnostics, and automation.
Practical familiarity with Kubernetes, Slurm, or equivalent cluster platforms. Enough to understand node onboarding, GPU scheduling, workload placement, networking, persistent storage, and day-2 operational requirements.
Experience working with technology vendors, delivery partners, customers, or cross-functional engineering teams.
Ability to turn incomplete or ambiguous requirements into explicit technical decisions, architecture principles, design artefacts, implementation plans, and measurable acceptance criteria.
Strong written and verbal communication skills, including the ability to explain technical trade-offs to engineering teams, vendors, customers, and senior decision-makers.
Comfort working in data-centre environments and reviewing physical infrastructure, server hardware, racks, cabling, power, cooling, and installation quality.
Experience with infrastructure automation or Infrastructure as Code, such as Ansible, Terraform, Python, Bash, GitOps, or equivalent tooling.
Fluent English and Polish is a must

Bonus experience:

Experience with NVIDIA AI infrastructure, including DGX, HGX, NVIDIA RTX PRO Servers, BasePOD, SuperPOD, DGX SuperPOD, NVIDIA AI Enterprise, or NVIDIA reference architectures.
Exposure to NVIDIA networking products, including Mellanox and ConnectX NICs, BlueField DPUs, Quantum InfiniBand switches, Spectrum Ethernet switches, LinkX cabling, NVIDIA UFM, or NVIDIA NEO.
Familiarity with AI platform components such as Kubernetes, Slurm, NVIDIA GPU Operator, DCGM, NVIDIA Base Command Manager, Run:ai, Ray, Kubeflow, MLflow, vLLM, Triton Inference Server, or similar technologies.
Experience designing or sizing high-performance storage environments, including parallel file systems, NVMe-over-Fabrics, Lustre, DDN, VAST Data, NetApp, Pure Storage or equivalent platforms.
Familiarity with GPU workload characteristics, distributed training, collective communication patterns, NCCL, inference serving, LLM fine-tuning, RAG pipelines, and data-pipeline bottlenecks.
Experience with liquid-cooled infrastructure, direct-to-chip cooling, high-density GPU racks, or data-centre design for AI and HPC workloads.
Hands-on experience with infrastructure validation or performance tools such as DCGM, NCCL tests, iperf3, perftest, ibdiagnet, fio, MLPerf, or vendor-specific diagnostic tooling.
Experience in consulting, technical presales, solution architecture, RFP delivery, or another customer-facing technical role.

Your future role

End-to-End AI Infrastructure architecture. Translate customer requirements—including AI training, fine-tuning, inference, RAG, simulation, data volumes, performance targets, availability objectives, and future growth—into complete architectures across compute, storage, networking, data-centre infrastructure, and platform software.
Solution definition and technical leadership. Lead technical discovery, requirements gathering, architecture workshops, solution reviews, and design decisions. Establish the target-state architecture, identify dependencies and risks, define technical standards, and provide a clear path from concept through implementation.
Bill of Materials development. Develop detailed, technically validated, and commercially viable BOMs covering GPU servers, CPU and management nodes, storage systems, InfiniBand or Ethernet fabrics, switches, NICs, DPUs, optics, cables, racks, PDUs, power infrastructure, cooling equipment, and relevant software, subscriptions, and support services.
Vendor architecture coordination. Work directly with server OEMs, GPU vendors, storage providers, networking vendors, system integrators, and specialist partners to validate solution configurations, confirm compatibility, align performance and resiliency assumptions, review lead times, compare options, and resolve technical dependencies.
Compute architecture and sizing. Define infrastructure for AI training, inference, fine-tuning, LLM Inference and data-processing workloads. Size GPU capacity, GPU memory, CPU, RAM, topology, local NVMe, host networking, management infrastructure, and expansion paths based on workload requirements.
Storage architecture. Define high-performance storage architectures appropriate for AI and GPU-intensive workloads. Cover capacity, throughput, IOPS, metadata performance, caching, tiering, data protection, backup, replication, and integration with Kubernetes, Slurm, and other cluster platforms.
Network Fabric architecture. Define, together with network specialists and technology vendors, high-performance InfiniBand and Ethernet/RoCE fabrics for GPU clusters. Validate topology, non-blocking or oversubscribed design targets, switch placement, host connectivity, optics, cabling, routing, quality-of-service requirements, congestion management, and operational management needs.
Data Centre and Facility integration. Act as the technical liaison between customers, colocation providers, data-centre operators, facilities teams, installation contractors, and technology partners. Define and validate rack layouts, floor-space requirements, structural constraints, power density, A/B feeds, PDU design, cooling capacity, airflow, liquid-cooling requirements, cable pathways, and commissioning plans.
Physical deployment assurance. You will not be expected to pull every cable or mount every server yourself, but you will own the correctness of the delivered design. You will define installation standards, review rack elevations and port maps, validate cabling plans, oversee data-centre teams and third-party contractors, and ensure physical implementation matches the approved architecture.
Platform Integration. Partner closely with platform engineering teams to ensure that physical infrastructure is ready for Kubernetes, Slurm, GPU scheduling, workload orchestration, multi-tenancy, identity and access integration, observability, security controls, node lifecycle management, and operational automation.
Design validation and technical acceptance. Define acceptance criteria and coordinate validation of the complete environment, including hardware health, GPU configuration, network connectivity and fabric health, storage performance, node configuration, benchmark execution, burn-in testing, operational documentation, and handover.
Customer, presales, and delivery support. Create architecture documents, high-level and low-level designs, BOMs, rack elevations, network diagrams, deployment plans, technical proposal content, and statements of work. Lead technical workshops, support RFPs and customer presentations, and provide technical leadership throughout the solution lifecycle.

Selection process

We kindly ask you to upload your CV in English.

Shortlisted candidates will be contacted for the interviewing process.

If your CV would match requirements, expect steps below:

* Quick HR call

* One on-site technical interview

* On-line interview with Hiring Manager

* Final decision.

About Deloitte

Deloitte to różnorodność ludzi, doświadczeń, branż i usług, w których je realizujemy - w 150 krajach na świecie. To wyzwania intelektualne, dobry start zawodowy, możliwości ciągłego rozwoju i zebrania cennych życiowych doświadczeń. Musisz zrobić pierwszy krok - postawić kropkę na końcu wysyłanego CV, a potem podpisywanej umowy. Deloitte to po prostu dobry wybór. I kropka.

About the Company

Deloitte

Audits companies and advises on tax and technology

Deloitte audits companies and advises businesses and governments on tax, technology, and risk, and it is one of the Big Four, the world's four largest accounting firms. In 1849 its founder became the first independent auditor of a public company, Britain's Great Western Railway. Its auditors check companies' financial reports for errors and fraud before investors rely on them, while its consultants design strategy, install business software, and defend clients against cyberattacks. Clients pay fees for each audit or project, and each country's Deloitte firm is a separate business owned by its local partners. Deloitte was founded in London in 1845 and is based in New York.

Company size: 10,001+
Company stage: Late Stage VC
Total funding: $17.1M
Headquarters: New York, NY
Founded: 1845

Recently Posted Jobs

Explore more opportunities at Deloitte

See more jobs