Director | AI / ML | Bengaluru | Engineering | Hybrid Cloud Engineering

Job description

Job requisition ID :: 107011 Date: Jul 3, 2026 Location: Bengaluru Designation: Director Entity: Deloitte Touche Tohmatsu India LLP **Your work profile** - AI Data Center Architecture & Solution Design - Design and implement AI-focused Data Center architectures aligned with Tier II, Tier III, and Tier IV standards. - Develop end-to-end AI Data Center solutions, including retrofitting traditional CPU-based data centers into AI Factories. - Create advisory documents, RFPs, technical proposals, and commercial proposals for AI Data Center engagements. - Design AI infrastructure solutions across hyperscalers (AWS, Azure, GCP, OCI) and NVIDIA Cloud Partners. - Prepare HLDs, LLDs, network diagrams, rack layouts, BOQs, and TCO models. - AI Networking & Fabric Architecture - Architect and deploy InfiniBand and NVIDIA Spectrum Ethernet fabrics for AI workloads. - Design and implement Spine-Leaf network architectures using EVPN-VXLAN overlays. - Configure and optimize BGP, ECMP, RoCE, and high-performance networking environments. - Lead Cumulus Linux-based deployments and network automation initiatives. - Optimize network performance, latency, throughput, and congestion management for AI environments. - AI Compute & GPU Infrastructure - Design and size GPU clusters using NVIDIA H100, H200, B200, B300, DGX, and AI Factory platforms. - Perform GPU capacity planning and workload profiling for AI and ML use cases. - Implement GPU virtualization and Multi-Instance GPU (MIG) architectures. - Support AI training and inference infrastructure deployments. - AI Storage & Platform Engineering - Design AI storage solutions utilizing NAS, SAN, NVMe, Object Storage, NFS, iSCSI, Fibre Channel, and parallel file systems. - Implement and manage Kubernetes-based AI platforms, including OpenShift and VMware Tanzu. - Deploy and integrate RUN and Slurm workload schedulers for GPU orchestration. - Ensure seamless integration of AI platforms with existing enterprise infrastructure. - Monitoring, Observability & Operations - Implement NVIDIA UFM, NVIDIA Mission Control, and NetQ for infrastructure monitoring and observability. - Configure telemetry, validation, troubleshooting, and fabric management workflows. - Drive infrastructure benchmarking, performance optimization, and capacity planning initiatives. - Support POCs, design validation exercises, production rollouts, and operational readiness activities. - Cloud & AI Services - Design AI infrastructure solutions across AWS, Azure, GCP, and OCI. - Enable AI services integration across hybrid and multi-cloud environments. - Provide guidance on AI platform adoption, scalability, and operational best practices. **Key skills required** **Data Center Infrastructure** - Strong understanding of Data Center power infrastructure, including UPS, PDU, ATS, switchgear, transformers, and generators. - Knowledge of Data Center cooling technologies such as CRAC, CRAH, liquid cooling, immersion cooling, and chiller systems. - Experience in rack design, cabling architecture, white space planning, and physical infrastructure design. - Understanding of raised floors, fire suppression systems, plenum design, and facility infrastructure. **AI Networking** - Strong expertise in InfiniBand (HDR/NDR), RoCE, and Ethernet fabrics. - Hands-on experience with NVIDIA Spectrum switches. - Deep understanding of EVPN-VXLAN, BGP, ECMP, Spine-Leaf architecture, and network automation. - Experience with Cumulus Linux environments. **AI Compute & Platforms** - Expertise in NVIDIA GPU platforms including DGX, H100, H200, B200, and B300. - Experience with GPU virtualization, MIG, and AI workload optimization. - Strong understanding of AI training and inference infrastructure. **AI Storage** - Knowledge of AI storage architectures and parallel file systems such as Lustre and GPFS. - Experience with NAS, SAN, Fibre Channel, NVMe, NFS, iSCSI, and Object Storage technologies. **Orchestration & Container Platforms** - Experience with Kubernetes ecosystems. - Hands-on expertise with OpenShift and VMware Tanzu. - Experience with RUN and Slurm workload management platforms. - Understanding of container networking for AI workloads. **AI Software Stack** - Understanding of AI infrastructure software layers including: - LLM Models - MLOps Platforms - Training and Inference Frameworks - Agentic AI - NVIDIA AI Enterprise - NVIDIA Licensing - NVIDIA NVIS **Cloud Technologies** - Strong understanding of AWS, Azure, GCP, and OCI services. - Experience designing AI and cloud-native solutions in hyperscaler environments.