The consultant will assess, design, and migrate on-premises computational workloads to AWS while optimizing cost governance and automation. They will also facilitate discovery workshops, manage machine image pipelines, and provide technical support during cutover and knowledge transfer.
Position Overview We are seeking an experienced AWS High-Performance Computing (HPC) Migration & Cloud Architecture Consultant to support the assessment, design, migration, and optimization of computational workloads from on-premises environments to AWS. This customer-facing role combines hands-on HPC and AWS expertise with architecture consulting, requirements discovery, cost governance, automation, and knowledge transfer. The consultant will help identify migration candidates, assess existing capabilities, design improvements, and support workload deployment in the customer’s target non-production AWS environment. The consultant will also assess machine image pipelines and the Open OnDemand user portal, facilitate discovery and design for analytics and AI agent integrations, and provide acceptance and cutover support. Key Responsibilities Workload Discovery & Requirements
Identify and assess on-premises computational workloads suitable for migration to the customer’s target AWS environment.
Develop and maintain an inventory of agreed-upon migration candidates, referred to as “In-Scope Workloads.”
Develop user personas and facilitate discovery workshops and technical deep-dive sessions.
Document operational requirements, data workflows, data requirements, and technical requirements for the In-Scope Workloads.
AWS Architecture & Workload Deployment
Review the initial baseline architecture for the customer’s target non-production AWS environment.
Advise and assist with deployment of the In-Scope Workloads.
Assess baseline deployments against AWS best practices and document recommendations.
Assist the customer in implementing approved optimization and remediation recommendations.
Cost Governance & Automation
Assess existing AWS cost governance capabilities, including tagging, cost allocation, anomaly detection, cost tracking, alerts, and reporting.
Review existing automated notifications and cost management processes.
Recommend improvements aligned with AWS best practices.
Design improved mechanisms for automated cost tracking, notification, management, and reporting.
Document cost tracking and reporting configurations.
Machine Image Pipeline & User Portal
Review the design and deployment of the existing automated Amazon Machine Image (AMI) pipeline.
Recommend improvements to image build, configuration, and deployment processes.
Assess the configuration and deployment of Open OnDemand, the customer’s open-source HPC web portal.
Recommend portal improvements and support integration with the AWS HPC environment, including authentication and user access requirements.
Document portal integration and configuration.
Analytics & AI Integration
Facilitate discovery sessions to identify analytics and AI agent integration opportunities that support the In-Scope Workloads.
Document integration requirements and proposed designs.
Acceptance, Cutover & Knowledge Transfer
Provide up to five business days of support to help troubleshoot and resolve issues during customer acceptance and cutover of the In-Scope Workloads.
Prepare technical documentation and handoff materials.
Conduct a Technical Exchange Workshop to transfer knowledge to the customer’s technical and operational teams.
Required Qualifications HPC & Compute
Hands-on experience migrating or operating HPC workloads on AWS using technologies such as AWS ParallelCluster, AWS Batch, and Amazon EC2 HPC instance families, including Hpc6a or Hpc7g.
Experience with placement groups and Elastic Fabric Adapter (EFA) for low-latency networking.
Experience with on-premises HPC environments and workload schedulers such as Slurm, PBS, LSF, or comparable platforms.
“I was the first applicant for a remote marketing position that got listed on the company website the same day I applied. Had an interview within 48 hours!”