New cohorts enrolling monthly — onsite, online & hybrid formats. Modern SQA Formula Bootcamp 2026 Registration ongoing Corporate teams: ask about custom bootcamp pricing.
AI testing team reviewing data quality model metrics subgroup results and drift
SQA, Test Automation & SDET · Course #058

AI/ML System Testing

AI/ML System Testing turns a specialised topic into a coherent, job-relevant workflow. Participants learn to test machine-learning systems across data, model, integration and operational lifecycle risks that differ from deterministic software, using ai/ml lifecycle and distinctive testing challenges to establish the foundation and intended use, quality characteristics and acceptance criteria to deepen the work. Learners examine data quality and leakage, select functional performance metrics, design subgroup and robustness tests, validate pipelines and monitor drift, failures and human-oversight controls. The final part connects data provenance, quality, representativeness and leakage with reporting, review and improvement so that the result can support an actual quality decision. The course connects statistical model evidence with system behaviour, user impact and repeatable TEVV documentation.

Enroll Interest Back to Courses

Overview

AI/ML System Testing turns a specialised topic into a coherent, job-relevant workflow. Participants learn to test machine-learning systems across data, model, integration and operational lifecycle risks that differ from deterministic software, using ai/ml lifecycle and distinctive testing challenges to establish the foundation and intended use, quality characteristics and acceptance criteria to deepen the work. Learners examine data quality and leakage, select functional performance metrics, design subgroup and robustness tests, validate pipelines and monitor drift, failures and human-oversight controls. The final part connects data provenance, quality, representativeness and leakage with reporting, review and improvement so that the result can support an actual quality decision. The course connects statistical model evidence with system behaviour, user impact and repeatable TEVV documentation.

Course Highlights

Lifecycle-based AI testing

Training and test-data quality

Model performance and uncertainty

Subgroup and robustness evaluation

Pipeline and system integration

Modules & Curriculum

AI/ML lifecycle and distinctive testing challenges

Intended use, quality characteristics and acceptance criteria

Data provenance, quality, representativeness and leakage

Train, validation and test-set integrity

Functional performance metrics and threshold trade-offs

Robustness, metamorphic and perturbation testing

Bias, subgroup performance and impact analysis

Model pipeline, API and application integration

Drift, monitoring, fallback and human oversight

TEVV evidence package and revalidation plan

Learning Outcomes

  • Identify AI-specific quality risks and acceptance criteria.
  • Test data completeness, validity, representativeness and leakage.
  • Select and interpret suitable model performance metrics.
  • Evaluate robustness and performance across relevant subgroups.
  • Validate model integration, fallback and human-oversight behaviour.
  • Design monitoring for drift, incidents and re-evaluation triggers.

Ready to Build Your Future-Ready Skills?

Test AI and ML systems across data, model, integration and operational risk.

Enroll Interest

Related Courses

Beginner Junior QA learners mapping assurance activities across a software delivery lifecycle

Software Quality Assurance (SQA) Fundamentals

View Course
Beginner Software team tracing requirements through development testing release and maintenance stages

Software Development Life Cycle (SDLC)

View Course
Beginner QA trainees organising test planning execution defect tracking and closure artefacts

Software Testing Life Cycle (STLC)

View Course