AI Infrastructure

GPU computing, distributed training, model serving, inference optimization, and AI platform engineering.

Topics

Lessons

RAPIDS, CUDA-enabled libraries, GPU ML pipelines

Jetson, Coral, on-device inference, model quantization

Feast, Tecton, SageMaker Feature Store — feature management

CUDA, NVIDIA A100/H100, AMD MI300, GPU clusters, vLLM

Drift detection, A/B testing, evaluation pipelines, guardrails

TensorRT, Triton, Ollama, text-generation-inference, ONNX