// structured learning

Production AI Engineering Curriculum

Unlike isolated talks, every IMLG meetup forms part of a long-term engineering curriculum. Each topic builds upon previous sessions.

TRACK 01

Linux for AI Infrastructure

Linux AdminNUMAHugePagescgroupsContainersOCIFilesystems

TRACK 02

GPU Architecture

CUDA CoresTensor CoresHBMWarp SchedulingStreaming MultiprocessorsNVLinkNVSwitch

TRACK 03

CUDA Software Stack

CUDA RuntimeCUDA DrivercuBLASNCCLcuDNNTensorRTTriton

TRACK 04

Distributed AI

PyTorchFSDPMegatronDeepSpeedPipeline ParallelismTensor ParallelismExpert Parallelism

TRACK 05

Inference Systems

vLLMSGLangTensorRT-LLMKV CacheSpeculative DecodingContinuous Batching

TRACK 06

MLOps

MLflowKubeflowKServeFeature StoresModel RegistryCI/CD

TRACK 07

LLMOps

RAGAgent SystemsEvaluationGuardrailsObservabilityPrompt Engineering

TRACK 08

Platform Engineering

KubernetesGPU OperatorsCiliumService MeshStorageNetworking

TRACK 09

Observability

OpenTelemetryDCGMPrometheusGrafanaClickHouseLokiTempo

TRACK 10

Production AI

End-to-End DeploymentEnterprise ScaleAI PlatformsArchitectureCase Studies
Attend a Meetup → Join the Community