What are the responsibilities and job description for the AI Infrastructure Director position at VinSmart Future?
Work Location: Hanoi: TechnoPark Tower, Vinhomes Ocean Park, Hanoi, Vietnam
Job overview:
We are investing in hyperscale AI infrastructure, with a long-term vision to develop multi-GW AI data center capacity and large-scale GPU compute clusters, enabling frontier model training and AI deployment at national scale.
We are looking for a Director of AI Infrastructure to architect, build, and scale our end-to-end AI infrastructure stack - from data center design to distributed training and inference systems.
This is a founding leadership role with full ownership of architecture, technology decisions, and ecosystem partnerships to build one of the largest AI infrastructure platforms in Southeast Asia.
What you will own:
1. End-to-End AI Infrastructure Architecture
Define and design the full stack of AI infrastructure:
→ Hyperscale data centers (multi-MW → GW-scale roadmap)
→ GPU clusters (hundreds → thousands GPUs, scaling toward 10,000 )
→ High-performance networking and distributed storage
Architect systems for frontier-scale model training and large-scale inference
2. Hyperscale Compute & Distributed Systems
- Build and operate large-scale GPU clusters (Phase 1: 500–1000 GPUs, scaling beyond)
- Design distributed systems for training, scheduling, and orchestration
- Optimize throughput, latency, utilization, and cost efficiency
3. AI Platform & Inference Systems
- Lead development of training and inference platforms
- Design systems for serving AI models at scale (millions of users / requests)
- Optimize model deployment, performance, and reliability
4. Data Center Strategy & Execution
- Lead design and scaling of AI-ready data centers (tens of MW → multi-GW roadmap)
- Drive decisions on power density, cooling (liquid/air), and rack design
- Own vendor ecosystem across NVIDIA, OEMs, networking, and infrastructure partners
5. Build the Team & Ecosystem
- Build and lead a world-class AI infrastructure organization
- Define engineering culture, hiring standards, and execution model
- Establish partnerships with global leaders in AI, cloud, and hardware ecosystems
What we are looking for:
We are open to candidates from different backgrounds, as long as you have built systems at scale in at least one of the following:
Option A - Distributed AI / Compute
- Built or operated large-scale distributed training systems
- Deep expertise in GPU clusters, HPC, or AI compute infrastructure
Option B - AI Platform / Inference
- Built AI platforms or large-scale inference systems
- Strong understanding of LLMs, training pipelines, and model optimization
Option C - Data Center / Hardware Infrastructure
- Designed or operated hyperscale data centers (MW-scale or beyond)
- Strong expertise in power, cooling, and high-density compute systems.
You might be a fit if you have:
- Experience working with hundreds → thousands → tens of thousands of GPUs
- Built systems for frontier AI / hyperscale cloud / HPC environments
- Strong hands-on technical depth combined with leadership and builder mindset
Why this role is unique
- Greenfield build - no legacy constraints
- Real scale commitment - multi-MW to 3–4GW AI infrastructure roadmap
- Full ownership - architecture, vendor selection, and execution
- Opportunity to shape national-scale AI capabilities
Compensation & Benefits
- Flexible executive compensation (structured based on candidate profile)
- Relocation & long-term incentives
- Access to Vingroup ecosystem (housing, healthcare, education…)
📩 Apply via:
Email: v.hanhtt41@vinsmartfuture.tech
For more information, please direct message me.