Skip to main content

Managing the AI Initiative Portfolio

Purpose​

This document describes how the AI initiative portfolio is maintained: the registry, statuses, distribution across stages, balance, and alignment with strategy and resources.

Key ideas​

  • Single registry — all initiatives (from idea to production) are tracked in one place with attributes: owner, stage, priority, target impact, risks.
  • Portfolio balance — a mix of quick wins and strategic initiatives; limiting the number of initiatives at the "Experiment" and "Prototype" stages based on resources.
  • Alignment with strategy — prioritization is aligned with business goals and the AI strategy; the portfolio is reviewed periodically.

How it works​

The AI Office maintains the registry (see data-model, initiative-entity). Portfolio management includes: adding/closing initiatives, assigning priorities, reporting on KPIs (portfolio-kpis), and stage gate decisions. Resources (teams, data, infrastructure) are planned with the portfolio in mind.

pie title Portfolio distribution by stage (example)
    "Idea" : 20
    "Experiment" : 25
    "Prototype" : 30
    "Production" : 15
    "Business Impact" : 10