AI Platform Reliability Engineer (PRE) /Architect

Location - Noida, Bengaluru, Chennai, Hyderabad and Pune.

Role

An AI PRE Engineer (L4) acts as a Platform Reliability Architect and Production Readiness Leader responsible for defining, governing, and scaling enterprise-grade AI/ML platforms. This role ensures AI systems are production-ready, resilient, observable, secure, compliant, and cost-efficient at scale, while driving standardization, automation, and platform-wide governance across AI engineering, SRE, DevOps, and MLOps disciplines. The L4 PRE acts as a strategic bridge between platform engineering, AI teams, and business leadership, ensuring reliability and scalability of AI solutions.

Responsibilities

Architecture & Production Readiness Governance

• Define and govern enterprise-wide production readiness frameworks across: Platform, data, model, application, and security layers
• Establish organization-wide PRE standards, policies, and maturity models
• Define and drive release certification governance (PRR frameworks) across all AI workloads
• Own production readiness lifecycle from design → deployment → scale

SLO/SLI Strategy & Reliability Engineering

• Define and govern: SLO/SLI/SLA frameworks for latency, availability, quality, safety, drift
• Error budget strategy at platform and business levels
• Establish reliability engineering models for AI systems
• Align SLOs with: Customer experience, business KPIs, and cost targets

Reference Architecture & Platform Design

• Define and publish: Enterprise AI reference architectures for: LLM applications, RAG pipelines, vector stores, agent frameworks
• Batch, real-time, and streaming inference systems
• Standardize: Platform architecture patterns across multi-cloud and hybrid environments
• Drive platform abstraction and reusable AI services
• Design and architect the NVIDIA Enterprise AI Software stack and deployment of GPU enabled Kubernetes clusters.

Deployment & Release Engineering Governance

• Define enterprise standards for: Safe deployment strategies (canary, shadow, blue-green, A/B testing)
• Prompt/model versioning, rollback, and release controls
• Govern: Release pipelines, CI/CD, GitOps, and policy-as-code frameworks
• Establish release risk assessment frameworks for AI workloads

Observability & AI Ops Strategy

• Define enterprise observability strategy covering: Metrics, logs, traces, and AI-specific telemetry: Token usage, latency, model drift, GPU utilization, quality metrics
• Standardize: Observability patterns for AI systems across platform
• Drive adoption of: AIOps and predictive reliability models
• Establish: SLO burn alerts, cost anomaly detection frameworks

Capacity Engineering & FinOps for AI

• Own: Capacity planning and forecasting models for: Token usage
• GPU/CPU scaling
• Vector databases and caching layers
• Define: Cost optimization strategies (FinOps for AI platforms)
• Govern: Resource allocation, autoscaling policies, and utilization efficiency

Resilience Engineering & System Design

• Define: Enterprise resilience patterns, including: Circuit breakers, fallbacks, retries, timeouts
• Semantic caching and prompt optimization
• Architect: High-availability AI systems with: Multi-region / multi-zone failover
• Establish: DR/BCP strategies with defined RTO/RPO targets

AI Security, Risk & Compliance Governance

• Define and govern: AI security frameworks: Secrets management
• Private networking
• Egress controls
• Model and prompt security
• Establish: Red-team validation and AI safety testing standards
• Own: Compliance frameworks mapping (ISO 27001, SOC 2, GDPR/DPDP, HIPAA)
• Ensure: Data privacy, auditability, and governance across AI platforms

Platform Engineering & Developer Enablement

• Define: Enterprise delivery frameworks: CI/CD pipelines
• SDKs, APIs, Helm charts, Terraform templates
• Drive: Platform standardization and self-service enablement
• Build: Reusable automation frameworks and PRE accelerators
• Lead: Enablement programs (trainings, brown-bags, playbooks)

Cross-Functional Leadership & Stakeholder Engagement

• Partner with: AI/ML teams, platform teams, DevOps, and business stakeholders
• Act as: Technical advisor to leadership and customers
• Validate: Architecture, NFRs (scale, latency, cost, safety) across programs
• Drive: Enterprise-wide adoption of PRE practices

Incident Management & Reliability Operations

• Own: Enterprise incident response frameworks for AI systems
• Lead: Critical incident command and escalations
• Govern: RCA, postmortems, and systemic reliability improvements
• Establish: Continuous improvement and

Qualifications & Experience

• Bachelor’s degree in Computer Science / Engineering (mandatory)
• Master’s preferred (Distributed Systems / Cloud / AI Systems)
• 12–18 years overall IT / platform engineering experience
• 7–10 years in AI / platform / cloud engineering at scale
• 5+ years in architecture / PRE / SRE strategy roles

Core Experience:

• Enterprise-scale:
• Production readiness frameworks, PRR governance
• SLO/SLI/SLA design and reliability engineering
• Deep expertise in:
• Kubernetes platforms (EKS/GKE/AKS) at scale
• GPU-aware AI infrastructure
• Multi-cluster, multi-tenant environments
• Strong experience in:
• MLOps / LLMOps / AI production systems
• Deployment strategies and release governance

Technology Experience:

• Cloud: AWS / Azure / GCP (architect-level understanding)
• Infra: Kubernetes, Docker, Terraform, GitOps
• Observability: Prometheus, Grafana, OpenTelemetry
• AI stack:
• LLM deployment, RAG pipelines, vector DBs
• Frameworks like LangChain, OpenAI
• NVIDIA AI Enterprise (NVAIE)
• Programming:
• Python / Go / Java (platform engineering focus)

Certifications Required

• NVIDIA Certified Professional – AI Infrastructure & Operations
• NVIDIA DLI – Advanced AI Infrastructure / GPU platforms
• Kubernetes Certifications (CKA / CKS – mandatory)
• Cloud Architect Certifications (AWS/Azure/GCP – Professional level preferred)
• DevOps / CI-CD certifications (preferred)
• Linux Certifications (RHCE / LFCS – advanced preferred)