Lead DevOps Engineer (AI & GPU) in भारत

This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Lead DevOps Engineer (AI & GPU) based in India (Bangalore, Remote). As a Lead DevOps Engineer (AI & GPU), you will lead the design, deployment, and operation of the infrastructure that powers production-grade AI services and high-performance inference workloads. You will build robust CI/CD pipelines, optimize GPU-based model serving, and establish reliable delivery processes across cloud, on-premises, edge, and air-gapped environments. Your work will ensure AI applications are deployed securely, operate efficiently, and remain observable throughout their lifecycle. You will oversee platform observability, resource utilization, cost attribution, and governance to deliver scalable, transparent, and economically sustainable AI operations. Collaborating with platform, security, data, and engineering teams, you will integrate automation, continuous verification, and policy enforcement into every stage of delivery. This role offers the opportunity to shape advanced AI infrastructure while solving complex challenges involving performance, security, reliability, and multi-tenant operations. Accountabilities: - Lead the design, implementation, and operation of CI/CD and continuous verification (CV) pipelines for AI services, agents, and runtime components, covering automated builds, testing, releases, validation, and rollback. - Implement and manage GitOps-based deployment workflows using tools such as Argo CD, Helm, GitHub Actions, GitLab CI, or equivalent technologies to ensure consistent and repeatable delivery. - Operate and optimize production-scale AI inference and model-serving infrastructure using frameworks such as Ray Serve, vLLM, NVIDIA Triton, or NVIDIA NIM across GPU-enabled environments.

- Manage AI asset governance and discovery, including artifact and container registries, model metadata, experiment tracking, upstream model repositories, software supply-chain security, license management, and lineage tracking. - Integrate tools such as Harbor, Artifactory, Nexus, MLflow, Hugging Face, NVIDIA GPU Cloud (NGC), Trivy, and software bill of materials (SBOM) scanning into secure AI delivery workflows. - Design and maintain a comprehensive observability platform encompassing metrics, logs, traces, dashboards, alerts, service-level agreements (SLAs), and AI-specific debugging and evaluation capabilities. - Build and operate OpenTelemetry collection infrastructure, including multi-tenant receivers, exporters, processing pipelines, batching, dynamic filtering, and Kafka-based telemetry delivery. - Implement GPU monitoring and performance visibility through tools such as the NVIDIA Data Center GPU Manager (DCGM) exporter, enabling accurate utilization tracking and infrastructure optimization. - Establish AI FinOps practices by implementing quotas, rate limits, utilization monitoring, and per-tenant cost attribution using tools such as OpenCost, Kubecost, or equivalent platforms. - Ensure that AI workload consumption, telemetry, and runtime behavior are consistently associated with tenant and workload identities to support accountability, cost control, and end-to-end traceability. - Integrate security, quality, integrity, compliance, and cost controls into deployment pipelines, ensuring releases meet defined policies before reaching production. - Connect observability and evaluation signals to deployment decisions, enabling progressive rollouts, automated rollback, and continuous improvements based on operational performance. - Support secure and consistent AI runtime operations across public cloud, private infrastructure, edge deployments, and air-gapped environments. - Collaborate with platform engineering, security, trust, and data teams to establish clear ownership boundaries, service expectations, consumption policies, and operational contracts.