CSC3119 · 9 modules

AI Model Deployment and Scaling

From notebook to hospital: pipelines, serialisation, APIs, cloud, Docker, canaries, ONNX, drift and XAI.

Start: Data Preparation, Preprocessing Leaks & Clinical Imbalance Mock exam Revision guide
0%
Exam readiness
Not started
  1. 1

    Data Preparation, Preprocessing Leaks & Clinical Imbalance

    Turn messy clinical hospital registries into leak-free deployable pipelines

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/2Lab 0/5
    Learn Lab
  2. 2

    Model Persistence: Pickle vs Joblib & Pipeline Packaging

    Serialize trained estimators, scalers, and encoders safely for zero-downtime serving

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  3. 3

    API Engineering with FastAPI, Pydantic & Async Workers

    Build ultra-low latency, strongly typed prediction endpoints with input contracts

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/2Lab 0/5
    Learn Lab
  4. 4

    Cloud Computing Concepts, Service Models & Hardware Sizing

    Understand NIST cloud characteristics, IaaS/PaaS/SaaS/MLaaS, and GPU/CPU cost modeling

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  5. 5

    Containerization, Multi-Stage Docker & Security Hardening

    Package microservices with lightweight base images, layer caching, and rootless users

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  6. 6

    Serving Architectures, Load Balancing & Canary Traffic Routing

    Implement zero-downtime upgrades, blue-green environments, and canary error budgets

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  7. 7

    Inference Performance: ONNX, Pruning & INT8 Quantization

    Accelerate prediction throughput, shrink model sizes, and reduce server operational costs

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  8. 8

    Observability, Covariate Drift & Population Stability Index (PSI)

    Detect data distribution changes and performance degradation before models fail patients

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/1Lab 0/4
    Learn Lab
  9. 9

    Capstone: Regional Referral Hospital Stroke Risk Pipeline & Explainable AI (XAI)

    Deploy the Ministry of Health stroke predictor with threshold calibration and SHAP feature attributions

    0%
    Lesson 0/64 note sectionsCheckpoint —Quiz 0/8Code 0/2Lab 0/4
    Learn Lab