About the Role

We are looking for a Systems or Solutions Architect with deep expertise in networking, infrastructure-as-a-service (IaaS), and cloud-scale system design to help architect and optimize AI/ML infrastructure.

The ideal candidate combines robust fundamentals in cloud architecture (AWS or equivalent), networking, compute, and storage, with hands-on experience in Kubernetes, observability, and automation.

You’ll design scalable systems that support large AI workloads — enabling productive training, inference, and data pipelines across distributed environments.

Key Responsibilities
Architect and scale AI/ML infrastructure across public cloud (AWS / Azure / GCP) and hybrid environments.
Design and optimize compute, storage, and network topologies for distributed training and inference clusters.
Build and manage containerized settings using Kubernetes, Docker, and Helm.
Develop automation frameworks for provisioning, scaling, and monitoring infrastructure using Python, Go, and IaC (Terraform / CloudFormation).




Partner with data science and ML Ops teams to align AI infrastructure requirements (GPU/CPU scaling, caching, throughput, latency).
Implement observability, logging, and tracing using Prometheus, Grafana, CloudWatch, or Open Telemetry.
Drive networking automation (BGP, routing, load balancing, VPNs, service meshes) using software-defined networking (SDN) and modern APIs.
Lead performance, reliability, and cost-optimization efforts for AI training and inference pipelines.
Collaborate cross-functionally with product, platform, and operations teams to ensure secure, performant, and resilient infrastructure.

Required Qualifications
Knowledge of AI/ML infrastructure patterns, including distributed training, inference pipelines, and GPU orchestration.
Bachelor’s or Master’s degree in Computer Science, Information Technology, or related field.
10+ years of experience in systems, infrastructure, or solutions architecture roles.
-

📌 Ai Infrastructure Systems/solutions Architect Gurugram
🏢 One moment, please
📍 Gurugram