Role:- Principal AI Platform Architect (Google Cloud Platform GKE | TPU/GPU Infrastructure)
Location:- Remote
Duration: Long-Term Contract
Job Title
Experience
12+ Years of Overall IT Experience with 5+ Years in AI/ML Platform Architecture on Google Cloud Platform (Google Cloud Platform)
Position Summary
Customer is seeking a highly experienced and hands-on Principal AI Platform Architect to lead the design and implementation of a next-generation AI infrastructure platform on Google Cloud Platform (Google Cloud Platform). The ideal candidate will possess deep expertise in architecting and deploying Google Kubernetes Engine (GKE) clusters supporting both TPU and GPU workloads for large-scale AI and Machine Learning applications.
This role requires a technical leader who can directly engage with engineering teams, drive architecture discussions, define platform standards, and establish a scalable, secure, and observable AI platform capable of supporting model development, training, experimentation, and production operations.
The successful candidate will have proven experience designing and implementing TPU-enabled clusters for large-scale model training while optimizing resource utilization, performance, cost, and operational efficiency.
Key Responsibilities
AI Platform Architecture & Design
GKE, TPU & GPU Infrastructure
AI Platform Operations & Engineering
Resource Allocation & Optimization
Observability & Monitoring
Security & Compliance
Stakeholder Engagement
Required Qualifications
Technical Skills Required
Google Cloud Platform
Kubernetes & Containerization
AI/ML Infrastructure
DevOps & Automation
Observability & Security
Preferred Qualifications
Success Criteria
The successful candidate will:
Key Skills Keywords
Google Cloud Platform, Google Kubernetes Engine (GKE), Cloud TPU, GPU Infrastructure, AI Platform Architecture, Kubernetes, TensorFlow, PyTorch, JAX, MLOps, Terraform, GitOps, Distributed Training, AI Infrastructure, Observability, Platform Engineering, Cloud Security, Capacity Planning, Resource Optimization, FinOps.
Writing Improvements