Set the overarching technical vision and strategy for compute infrastructure supporting large-scale AI/ML systems, establishing the direction that senior and lead architects build upon. 

Own the most complex, high-stakes architecture decisions across compute, networking, storage, orchestration, and model serving, rationalizing multiple viable options and making authoritative choices aligned to client situations, standards, and strategic objectives. 

Architect and hands-on prototype large-scale, cost-optimized compute and distributed training systems — building reference implementations, proofs-of-concept, and benchmarks to validate designs before they scale. 

Define reference architectures, standards, and architectural patterns that scale across engagements, and personally implement the foundational tooling, infrastructure-as-code, and automation that anchor them. 

Lead enterprise-scale architecture assessments and design reviews, getting hands-on in the environment to validate findings, profile real workloads, and demonstrate optimization opportunities. 

Shape and steward the AI infrastructure roadmap and technology strategy, planning capacity, scaling, and technology evolution in step with long-term business goals. 

Identify, evaluate, and hands-on pilot promising emerging technologies and innovations, building and testing them in real conditions to judge where and when they belong in a solution. 

Drive hands-on performance and cost optimization of the computational stack, profiling GPU/compute utilization, tuning distributed training and model-serving workloads, and engineering improvements to meet SLAs while controlling cost. 

Serve as the principal authority across hyperscaler cloud platforms, with current, hands-on expertise in AI/ML services, accelerators, interconnects, and cost levers, and breadth across multiple providers. 

Lead deep, hands-on troubleshooting and root-cause analysis of the most complex issues across the stack — hardware, networking, software, and models — resolving the problems others cannot and codifying the fixes. 

Cultivate and lead relationships with infrastructure partners and partnering organizations, securing early access, influence, and insight, and representing the practice's technical credibility in those forums. 

Provide executive- and client-level technical advisory, translating complex infrastructure trade-offs into clear, defensible recommendations tied to business outcomes. 

Define monitoring, observability, and reliability strategy across InfraOps and MLOps, and implement the instrumentation, SLAs, SLOs, and cost/performance governance for production AI/ML systems. 

Ensure enterprise integration, security, compliance, and regulatory alignment of AI/ML infrastructure across the firm's solutions. 

Mentor, elevate, and grow the architect community, developing senior and lead architects through hands-on pairing, design collaboration, and code/architecture reviews. 

Champion cost-efficiency and value realization, ensuring infrastructure not only performs and scales but demonstrably moves the needle for the organizations it serves.