AI Solutions Architect

Position Summary

The AI Solutions Architect is responsible for designing, positioning, and guiding the implementation of enterprise AI solutions built around NVIDIA AI technologies, DGX SuperPOD reference architectures, and the NVIDIA technology ecosystem.

This role bridges the gap between customer AI use cases and the infrastructure, platforms, software, and services required to deliver them.

The AI Solutions Architect works directly with customers to understand business objectives, AI workloads, data requirements, performance expectations, and operational constraints, translating those requirements into scalable AI solutions incorporating GPU compute, networking, storage, AI software platforms, data center infrastructure, cloud services, security, and operational services.

The role serves as a technical advisor to customers and a technical leader within the organization's AI practice.

Key Responsibilities

AI Solution Architecture

Lead the architecture of enterprise AI solutions supporting:

• Generative AI
• Large Language Models (LLMs)
• Retrieval-Augmented Generation (RAG)
• AI model training
• Model fine-tuning
• AI inference
• Agentic AI
• Data science
• Machine learning
• Computer vision
• High-Performance Computing (HPC)

Translate customer AI use cases into technical requirements across compute, GPU capacity, networking, storage, data, security, software, facilities, and operations.

Determine the appropriate deployment model including:

• Customer-owned AI infrastructure
• NVIDIA DGX-based infrastructure
• NVIDIA DGX SuperPOD
• NVIDIA-certified OEM infrastructure
• Colocation
• Private cloud
• Public cloud
• GPU-as-a-Service
• Hybrid AI environments

AI Architecture

Develop AI solutions aligned with NVIDIA reference architectures and validated designs.

Architect solutions incorporating technologies such as or equivalent to:

• NVIDIA DGX
• NVIDIA DGX SuperPOD
• NVIDIA HGX
• NVIDIA GB200/GB300 NVL systems
• NVIDIA H100/H200 and Blackwell-generation GPUs
• NVIDIA NVLink and NVLink Switch
• NVIDIA Spectrum-X
• NVIDIA Quantum InfiniBand
• NVIDIA BlueField DPUs
• NVIDIA ConnectX
• NVIDIA AI Enterprise
• NVIDIA NIM
• NVIDIA CUDA
• NVIDIA NCCL
• NVIDIA Base Command
• NVIDIA GPU Operator
• NVIDIA Network Operator

Maintain awareness of the industries evolving AI Factory solutions, architectures and product roadmaps.

AI Use Case Discovery

Conduct customer workshops to understand business and technical requirements for AI initiatives.

Work with business leaders, infrastructure teams, data scientists, application teams, security teams, and AI/ML engineers to identify:

• Business objectives
• AI use cases
• Model requirements
• Model sizes
• Training requirements
• Inference requirements
• Dataset sizes
• Data sources
• RAG requirements
• Performance requirements
• Security and governance requirements
• Availability requirements
• Growth expectations

Translate use cases into an actionable AI architecture and technology roadmap.

AI Infrastructure Architecture

Design the underlying infrastructure required to support enterprise AI workloads.

GPU Compute

Design NVIDIA GPU clusters ranging from smaller departmental environments through large-scale AI Factory and DGX SuperPOD deployments.

Networking

Design high-performance AI network fabrics incorporating the following or their industry equivalents:

• NVIDIA Spectrum-X Ethernet
• NVIDIA Quantum InfiniBand
• RDMA/RoCE
• NVIDIA BlueField
• NVIDIA ConnectX
• GPUDirect
• High-performance Ethernet
• Leaf/spine architectures

Storage

Design AI storage platforms optimized for:

• Training datasets
• Model repositories
• Checkpointing
• RAG data
• High-performance parallel access
• AI data pipelines
• Long-term data retention

Evaluate NVIDIA-certified storage solutions from vendors such as Dell Technologies, NetApp, Pure Storage, VAST Data, WEKA, DDN, IBM, HPE, and other approved ecosystem partners.

AI Platform Architecture

Design the platform layer required to operate AI workloads at enterprise scale.

Solutions may incorporate:

• NVIDIA AI Enterprise
• NVIDIA NIM
• Kubernetes
• Red Hat OpenShift
• Slurm
• Containerized GPU environments
• MLOps platforms
• Model repositories
• API gateways
• Vector databases
• Data pipelines
• AI observability
• Infrastructure automation

Work with AI development teams to ensure infrastructure and platforms are optimized for the intended workloads.

Data & RAG Architecture

Assist customers in designing the data architecture supporting enterprise AI.

Provide architectural guidance around:

• Enterprise data sources
• Data ingestion
• Data pipelines
• Object and file storage
• Vector databases
• Embedding pipelines
• RAG architectures
• Model access to enterprise data
• Data security
• Data lifecycle management

Partner with application, data, and AI specialists where deeper application-level expertise is required.

AI Security & Governance

Incorporate security into AI solution architectures.

Address requirements including:

• Identity and access management
• Data security
• Network segmentation
• API security
• Model access
• Data privacy
• AI workload isolation
• Logging and monitoring
• AI governance
• Model governance

Work with cybersecurity architects to integrate AI environments into the customer's broader enterprise security architecture.

AI Factory & Data Center Architecture

Assess whether customer facilities can support high-density AI infrastructure.

Provide architectural guidance for:

• Rack density
• Power
• Cooling
• Liquid cooling
• Power distribution
• Network cabling
• Fiber infrastructure
• Rack layouts
• Resiliency
• Physical security

Identify when customer requirements are better served through colocation, hosted infrastructure, cloud, or GPU-as-a-Service alternatives.

Technology Partner Ecosystem

Maintain strong technical relationships with key AI infrastructure partners.

Work with NVIDIA-certified and approved ecosystem vendors across:

• Compute
• Networking
• Storage
• Data center infrastructure
• AI software
• Data management
• Security
• Cloud
• Colocation
• Managed services

Evaluate new technologies and determine their appropriate role within the organization's AI solutions portfolio.

Participate in vendor roadmap sessions, NVIDIA technical briefings, solution validation, partner enablement, and Technical Review Boards.

Presales & Customer Engagement

Serve as the technical lead for strategic AI opportunities.

Responsibilities include:

• Customer discovery
• AI strategy workshops
• Technical presentations
• Whiteboarding
• Architecture development
• Solution sizing
• Bill of Materials development
• Proof-of-concept planning
• RFP/RFI responses
• Technical proposals
• Statements of Work
• Executive briefings
• Technical demonstrations

Communicate complex AI concepts effectively to audiences ranging from engineers and data scientists to CIOs and business executives.

Architecture Documentation

Produce professional customer-facing deliverables including:

• AI Solution Architectures
• High-Level Designs
• Logical architecture diagrams
• Physical architecture diagrams
• AI Factory designs
• Network topology diagrams
• Storage architectures
• GPU cluster architectures
• Bills of Materials
• Capacity models
• AI readiness assessments
• Implementation roadmaps
• Architecture Decision Documents

Ensure solutions can transition successfully from presales architecture into implementation.

Required Experience

• 8+ years of experience in enterprise infrastructure, cloud, data center, HPC, or solution architecture
• Experience designing AI, GPU, HPC, or accelerated-computing environments
• Strong knowledge of NVIDIA technologies and GPU architectures
• Understanding of generative AI, LLM, RAG, training, fine-tuning, and inference workloads
• Strong knowledge of enterprise compute, networking, and storage
• Experience with high-performance Ethernet and/or InfiniBand
• Understanding of Kubernetes and containerized environments
• Knowledge of Linux infrastructure
• Experience designing enterprise architectures and technical solutions
• Strong customer-facing consulting and presentation skills
• Experience working with enterprise OEMs and technology partners

Preferred Technical Experience

Experience across several of the following technology areas is highly desirable:

NVIDIA: DGX, DGX SuperPOD, HGX, GB200/GB300 NVL, H100/H200, Blackwell, Spectrum-X, Quantum InfiniBand, BlueField, ConnectX, NVLink, CUDA, NCCL, AI Enterprise, NIM and GPUDirect.

Compute: Dell Technologies, HPE, Lenovo, Supermicro and NVIDIA-certified compute platforms.

Networking: NVIDIA Networking, Arista, Cisco and high-performance Ethernet/InfiniBand.

Storage: NetApp, Pure Storage, Dell, VAST Data, WEKA, DDN, IBM and NVIDIA-certified AI storage platforms.

Platforms: Kubernetes, Red Hat OpenShift, Slurm, Linux and container platforms.

AI/Data: RAG architectures, vector databases, data pipelines, MLOps, model repositories and enterprise AI platforms.

Automation: Ansible, Terraform and infrastructure orchestration technologies.

Preferred Certifications

Relevant certifications may include:

• NVIDIA AI certifications
• NVIDIA AI Infrastructure certifications
• NVIDIA networking certifications
• Kubernetes certifications
• Red Hat certifications
• Cisco CCNP/CCIE
• Dell, DDN, Arista, NetApp, Pure Storage, or other relevant OEM certifications

Equivalent hands-on AI architecture experience may substitute for formal certifications.

Equal Opportunity

Insight Investments, Corp is an Equal Employment Opportunity/Affirmative Action employer and all qualified applicants will receive consideration for employment without regard to race, color, religion, national origin, gender, age, sexual orientation, gender identity, disability, status as a Veteran or other protected classification.