Principal Infrastructure Engineer

Arkana Laboratories•Arkansas, AR

19h•Onsite

About The Position

Arkana is the nation's leading renal pathology practice. We hold one of the richest kidney biopsy datasets in the world, and we are building the infrastructure to train clinical AI models on it. That includes an NVIDIA Blackwell class GPU cluster that almost no one in healthcare, and no one else in Central Arkansas, can match. It sits alongside a hybrid cloud platform, a laboratory information system built in house, and clinical AI products used by nephrologists every day. You will design and build the infrastructure behind all of it. The Principal Infrastructure Engineer is the senior most technical individual contributor in our technology operations organization. This is a pure technical leadership role with no direct reports. You own the architecture, build alongside the team, solve the hardest problems, and raise the bar for the engineers around you. You report to the Director of Technology Operations and work alongside a small team of senior infrastructure engineers. We created this role because the work demands it. The AI infrastructure buildout, the hybrid cloud migration, and the clinical platform architecture each need someone whose full time job is getting the design right and seeing it through to running systems.

Requirements

8+ years of deep infrastructure and systems engineering experience; there is no upper bound, and we welcome deeply experienced candidates
Expert level cloud platform knowledge, with deep Azure experience strongly preferred and AWS familiarity a plus, alongside genuine comfort with physical infrastructure: servers, storage, networking, power, and cooling
A track record of owning architecture, not just operating it: designs you can point to, trade offs you can defend
Hands on mastery of Infrastructure as Code (Terraform or similar), containerization, and modern automation
Experience in healthcare or another regulated environment, or the judgment to operate in one

Nice To Haves

GPU cluster, HPC, or ML infrastructure experience
Experience self hosting or serving large models (vLLM, Triton, or similar)
Storage and networking design for high throughput imaging or scientific workloads
Relevant certifications preferred: Azure Solutions Architect Expert, AWS Solutions Architect Professional, NVIDIA-Certified Professional, or equivalent

Responsibilities

Design and stand up our Blackwell class GPU training cluster, including node architecture, power and cooling requirements, high throughput NVMe storage for whole slide imaging workloads, and expansion planning for additional nodes
Architect inference infrastructure for self hosted open weight models serving internal and clinical applications
Design the data pipelines and storage tiers that feed model training on terabyte scale pathology imagery
Own cloud architecture for our custom applications, primarily on Azure with some AWS, including Arkana Connect and our analytics platforms
Design the target state hybrid model and the migration path from legacy on premises systems, keeping workloads where they perform and comply best
Set Infrastructure as Code, networking, and automation standards for the team
Serve as the escalation point for the most complex infrastructure problems
Mentor and raise the technical ceiling of the infrastructure team through design reviews, pairing, and standards
Advise the Director of Technology Operations on technology selection, vendor evaluation, and roadmap
Architect infrastructure that meets HIPAA and HITRUST requirements as a first class constraint, working closely with our InfoSec practice
Design backup, disaster recovery, and business continuity architecture for clinical operations