Skip to main content

Who this track is for

The ML Engineer track is for engineers who want to own the full lifecycle of ML systems — from training runs to production serving. You’ll go deep on PyTorch, learn how to scale training across GPUs, and build the infrastructure that keeps models healthy in production. Best fit if you:
  • Come from a software or data engineering background
  • Want to work on ML infrastructure at companies with serious model training
  • Are interested in the systems and engineering side of ML

Curriculum

Level 1 — Deep Learning Foundations (free)

ModuleKey Topics
Neural Networks & PyTorchTensors, autograd, training loops, GPU basics
CNN & RNN ArchitecturesConv layers, sequence models, practical applications

Level 2 — Training at Scale

ModuleKey Topics
Advanced PyTorchCustom datasets, DataLoaders, mixed precision, profiling
Distributed TrainingDDP, model parallelism, gradient accumulation

Level 3 — MLOps

ModuleKey Topics
Experiment TrackingMLflow, Weights & Biases, hyperparameter tuning
Model Registry & CI/CDVersioning, automated retraining, model promotion
Data Pipelines for MLFeature stores, data versioning, lineage

Level 4 — Model Serving

ModuleKey Topics
Inference OptimisationQuantisation, pruning, ONNX, TensorRT
Serving SystemsBentoML, Triton, latency budgets, autoscaling
Production MonitoringShadow deployment, A/B testing models, drift

Level 5 — Capstone

Design and build a complete ML system: data → training → serving → monitoring. Reviewed by your cohort on system design and production readiness.