Lead AI Platform Integrant Inc

The Lead AI Platform Engineer is responsible for bridging AI workloads with production-grade infrastructure, with a strong focus on NVIDIA AI stack, enabling high-performance, scalable, and optimized AI systems.

This role focuses on model optimization, runtime efficiency, and GPU utilization, ensuring that AI workloads are production-ready, cost-efficient, and performant across enterprise environments.

Roles and Responsibilities:

• Translate AI/ML workloads into optimized infrastructure and deployment strategies
• Optimize model performance across GPU environments (latency, throughput, memory utilization)
• Design and implement inference and training pipelines using NVIDIA stack tools (TensorRT, Triton, NIM)
• Convert and optimize models across frameworks (PyTorch ONNX TensorRT)
• Analyze and resolve performance bottlenecks using profiling tools (GPU, memory, network)
• Improve GPU utilization and scheduling efficiency across clusters
• Design scalable distributed training and inference architectures
• Work closely with customers to define AI infrastructure strategies and deployment models
• Support production deployments including monitoring, rollback, and performance validation
• Conduct applied research to improve model efficiency and infrastructure utilization
• Mentor team members on AI infrastructure, optimization, and GPU systems
• Experiment tracking tools (MLflow, W&B, Neptune) log parameters, metrics, and artifacts for comparison
• Find the Model degradation happens post-deployment: concept drift, data pipeline changes, traffic pattern shifts
• Root cause analysis (RCA) applies to ML systems: isolating variables, reproducing issues

Nice to Have

• Experience with NVIDIA NIM and NGC ecosystem
• Exposure to Megatron-LM, NeMo, or large-scale LLM training/inference
• Experience with LLM optimization techniques (KV cache, batching strategies)
• Familiarity with MLOps practices and CI/CD for AI systems
• Experience in customer-facing architecture or consulting roles
• Familiarity with hybrid cloud / on-prem HPC environments