Solution Architect- AI

Solution Architect AIDepartment: Cloud Presales Location: Mumbai / ChennaiJob PurposeTo architect, design, and implement GPU-enabled High Performance Computing (HPC) and AI/ML platform solutions that are scalable, secure, and optimized to support enterprise AI, ML/DL workloads, simulation, and large-scale data analytics. This role will define infrastructure strategies, workload placement, performance tuning, and managed service roadmaps for GPU and HPC platforms within the Data Centre business.Key ResponsibilitiesPlatform Architecture & Design- Architect GPU and HPC infrastructure platforms for AI/ML training, inference, and HPC workloads.- Design GPU-as-a-Service (GPUaaS) models including on-demand, reserved, and burst GPU clusters.- Integrate AI/ML frameworks (TensorFlow, PyTorch, KubeFlow, JupyterHub, etc.) into enterprise-ready stacks.Infrastructure & Workload Optimization- Optimize performance tuning, resource scheduling, and workload orchestration across HPC clusters and GPU nodes.- Enhance distributed training, model parallelism, and storage bandwidth utilization (NVMe, Lustre, GPFS, Ceph).AI/ML Platform Enablement- Provide cloud-native environments with containerized ML workflows (Kubernetes, Docker, Singularity).- Build and manage model hosting and inference platforms (REST APIs, containerized inference servers).Security & Compliance- Implement data security, encryption, access control, and compliance frameworks for sensitive AI/HPC workloads.- Architect air-gapped solutions for government/defense workloads when required.Technology Integration & Innovation- Evaluate and integrate next-generation GPUs (NVIDIA H200/A100/L40S, AMD MI300, etc.), HPC accelerators, and AI chipsets.- Enable hybrid and hyperconverged AI infrastructure combining GPU, CPU, and storage resources.Customer & Business Enablement- Collaborate with data scientists, researchers, and enterprise customers to align platform capabilities with business outcomes.- Define GPU/HPC platform services catalog and managed service offerings.Automation & DevOps- Implement MLOps pipelines, infrastructure as code (Terraform, Ansible), and workload scheduling (SLURM, Kubernetes).Qualifications & ExperienceEducational Qualifications- BE/B-Tech or equivalent in Computer Science, Electronics & Communication, or related fields.Experience- 8–12 years of overall IT experience, including 5+ years in HPC/AI/ML/GPU platform architecture.Technical Expertise- Strong knowledge of GPU architecture (NVIDIA, AMD) and HPC systems.- Proficiency with AI/ML frameworks such as TensorFlow, PyTorch, Keras, MXNet, Hugging Face.- Experience with distributed training and orchestration frameworks like KubeFlow, MLflow, Ray, Horovod.- Knowledge of parallel computing, MPI, CUDA, ROCm, and GPU drivers.- Familiarity with storage technologies such as NVMe, Lustre, GPFS, Ceph, and Object Storage for HPC/AI workloads.- Hands-on experience with GPU cloud platforms (AWS Sagemaker, Azure ML, GCP Vertex AI) and on-prem HPC cluster management.- Automation and MLOps expertise: CI/CD pipelines for ML, infrastructure as code, and workflow automation.- Understanding of security and governance including data privacy laws (e.g., DPDP Act), ISO, PCI-DSS, HIPAA compliance, and secure GPU cluster design.Certifications (Preferred)- NVIDIA Certified AI Specialist- Azure AI Engineer- AWS ML Specialty- HPC-related certificationsSoft Skills- Strong stakeholder communication and collaboration skills.- Ability to work effectively with data scientists, researchers, and enterprise IT teams.- Align technical solutions to business objectives with a strategic mindset.