Role Purpose
Enable enterprise customers to operationalize AI workloads by deploying and optimizing model-serving platforms (e.g., NVIDIA Triton, vLLM, KServe) within Rackspace’s Private Cloud and Hybrid environments. This role bridges AI engineering and platform operations, ensuring secure, scalable, and cost-efficient inference services.
Key Responsibilities : -
Model Deployment & Optimization
Package and deploy ML/LLM models on Triton, vLLM, or KServe within Kubernetes clusters.
Tune performance (batching, KV-cache, TensorRT optimizations) for latency and throughput SLAs.
Platform Integration
Work with VMware VCF9, NSX-T, and vSAN ESA to ensure GPU resource allocation and multi-tenancy.
Implement RBAC, encryption, and compliance controls for sovereign/private cloud customers.
API & Service Enablement
Integrate models with Rackspace’s Unified Inference API and API Gateway for multi-tenant routing.
Support RAG and agentic workflows by connecting to vector databases and context stores.
Observability & FinOps
Configure telemetry for GPU utilization, request tracing, and error monitoring.
Collaborate with FinOps to enable usage metering and chargeback reporting.
Customer Engineering Support
Assist solution architects in onboarding customers, creating reference patterns for BFSI, Healthcare, and other verticals.
Provide troubleshooting and performance benchmarking guidance.
Continuous Improvement
Stay current with emerging model-serving frameworks and GPU acceleration techniques.
Contribute to reusable Helm charts, operators, and automation scripts.
