Certified Professional in Managing AI - PMI-CPMAI™ - $289

Format- Hybrid- Both In-Person and Online Option to Attend Live Lectures

In-Person Location: 5800 Ambler Drive suite 210, Mississauga, Ontario L4W 4J4

Certification at a glance

Credential

PMI Certified Professional in Managing AI (PMI-CPMAI)™

Prior experience

No professional experience prerequisite.

Exam

120 total questions (100 scored + 20 unscored); 160 minutes

Course audience, prerequisites and outcomes

Audience: project/program/portfolio managers, product owners, Agile coaches, business analysts, data/AI initiative managers, transformation leads, governance and risk professionals, and sponsors of enterprise AI initiatives.

Suggested entry knowledge: basic project lifecycle, business-case analysis, delivery governance and familiarity with AI terms. No programming knowledge is required. Technical demonstrations can be done via supplied sample outputs.

By the end of this workshop, participants should be able to:

  • Frame an AI opportunity as a measurable business problem; determine whether AI is warranted and define success criteria.
  • Use the six phases of the CPMAI methodology as an iterative delivery framework.
  • Assess data sources, quality, privacy, security, provenance, bias and readiness for model development.
  • Coordinate model experimentation, acceptance testing and evidence-based go/no-go decisions.
  • Plan responsible AI controls, compliance documentation, human oversight and operational monitoring.

Build an actionable AI project delivery plan and interpret scenario-based certification questions.

4-day schedule

Detailed curriculum and learning activities

Module 1 | AI projects and CPMAI fundamentals (3 hours)

  • AI, machine learning, deep learning, generative AI, predictive systems and intelligent automation; practical differences.
  • Why AI initiatives fail: unclear value, unsuitable data, experimentation uncertainty, model drift, integration and adoption.
  • CPMAI as vendor-neutral, data-centric and iterative; six phases, phase gates and iterative feedback loops.
  • Project governance: sponsor, product owner, data scientist, data engineer, model risk, information security and business process owner.

Application: Exercise: AI-or-not-AI decision matrix for five business use cases.

Module 2 | Phase I: Business understanding (4 hours)

  • Stakeholder identification, process/problem discovery, personas and decision mapping.
  • Define baseline performance, value hypothesis, success measures, benefits and ROI/TCO.
  • Feasibility, build/buy/partner alternatives, constraints, risks and responsible-use screening.
  • AI project charter, scope, assumptions, data dependencies, acceptance criteria and decision rights.

Application: Deliverable: AI opportunity canvas, business case and initial AI project charter.

Module 3 | Phase II: Data understanding (4 hours)

  • Data categories, structured/unstructured sources, ownership, lineage and data discovery.
  • Assess availability, accessibility, completeness, representativeness, timeliness and suitability.
  • Training, validation and test partitions; leakage, sampling bias, label quality and provenance.
  • Privacy impact, access control, classification, retention, legal rights and third-party data risk.

Application: Exercise: Data inventory and suitability assessment with go/no-go recommendations.

Module 4 | Phase III: Data preparation (3 hours)

  • Data cleaning, feature selection/engineering, normalization, labeling and augmentation.
  • Pipeline orchestration, reproducibility, versioning, data quality checks and DataOps.
  • Mitigating bias, securing datasets, preparing retrieval corpora for generative AI and quality thresholds.
  • Data readiness review and documented transition into model experimentation.

Application: Deliverable: Data preparation backlog, quality dashboard and readiness checklist.

Module 5 | Phase IV: Iterative model development (4 hours)

  • Match use cases to supervised/unsupervised learning, LLMs, retrieval-augmented generation and automation.
  • Experiment design, baseline model, selection trade-offs, compute costs and vendor evaluation.
  • Model training, tuning, hyperparameters, evaluation datasets, prompt/version management and reproducible experiments.
  • Agile delivery: sprint experiments, hypothesis backlog, model cards, increment review and technical debt.

Application: Exercise: Model approach comparison and two-iteration experiment plan.

Module 6 | Phase V: Model testing and evaluation (3 hours)

  • Metric selection: precision, recall, F1, false-positive cost, calibration and business KPIs.
  • Generative-AI measures: groundedness, hallucination rate, answer quality, toxicity and human review.
  • Fairness and explainability checks; robustness, security, adversarial testing and red teaming.
  • Acceptance thresholds, validation evidence, risk exceptions and deployment decision.

Application: Deliverable: AI evaluation scorecard, test strategy and release recommendation.

Module 7 | Phase VI: Operationalization and adoption (4 hours)

  • Integration, deployment approaches, MLOps/LLMOps, model registry and release controls.
  • Monitoring for drift, latency, accuracy, cost, safety, bias and actual business outcomes.
  • Human-in-the-loop workflows, incident response, rollback, retraining triggers and model retirement.
  • Change management, training, stakeholder communication, policy updates and benefits realization.

Application: Exercise: Operating model, KPI monitoring plan and adoption roadmap.

Module 8 | Integrated governance, capstone and exam preparation (3 hours)

  • Cross-cutting responsible AI: privacy, security, ethics, transparency, accountability and auditability.
  • Map project artefacts to PMI’s five current exam domains and their relative weightings.
  • Scenario-based question drills, distractor analysis and personal study strategy.
  • Team capstone presentation: defend investment, data readiness, testing and operational readiness.

Application: Deliverable: Integrated AI project playbook and individual exam study plan.