Case Study 3: AI Knowledge Portal

This case study presents the development of an enterprise AI enablement product at The World Bank, designed to support employees across different stages of AI adoption, from discovering AI capabilities and building knowledge to using AI responsibly and exploring opportunities to build AI solutions. The AI Knowledge Portal brings together AI products, learning, community, governance guidance, and development resources into a single experience supporting organisation-wide AI adoption.

PRODUCT STRATEGY
Product Context
The AI Knowledge Portal serves as the Enablement layer of the World Bank Group AI Platform, following the Access layer with AI Chat and AI Search.
Its role is to help employees navigate the growing AI landscape, from discovering available products and resources to learning, engaging with the AI community, using AI responsibly, and eventually building AI solutions.
Business Problem
As AI capabilities expanded across the organisation, the World Bank Group faced a growing challenge of making these capabilities discoverable, understandable, and scalable across a diverse global workforce.

Key organisational challenges:
  • connecting AI access with broader adoption;
  • increasing visibility and reuse of existing AI capabilities;
  • connecting learning and guidance with practical application;
  • scaling responsible AI adoption;
  • creating pathways from AI use to experimentation and development.
Product Vision
Create central AI Knowledge Portal that helps employees move from discovering AI to learning, engaging, using AI responsibly, innovating, and ultimately building with it. The Portal connects AI products, education, community, governance, innovation, and development resources into a coherent experience that supports organisation-wide AI adoption.
AI Use Case Discovery
The product scope was shaped at the intersection of employee needs and organisational priorities, with analysis of AI adoption journeys helping identify the key product capabilities required to support adoption.

These translated into core product areas:
  • Discover. Find internal AI initiatives and approved external AI capabilities.
  • Learn. Build AI knowledge and practical skills, and learn from AI Champions.
  • Engage. Connect with AI communities, events, and organisational knowledge.
  • Use Responsibly. Access governance, policies, and responsible AI guidance.
  • Innovate. Explore existing AI ideas and propose new opportunities for AI solutions.
  • Build. Access AI Factory and approved APIs to safely experiment with AI capabilities and evaluate their potential for new internal AI initiatives.
PRODUCT DISCOVERY
User Centered Discovery
Product discovery was grounded in organisational priorities and focused on understanding the employees the Portal needed to support. Key employee segments and their business objectives helped define the AI needs the product needed to address, while key user journeys showed how employees would engage with AI across different scenarios. These insights informed the product capabilities and architecture that followed.

Key Employee Segments / User Personas

  • Data & Research
    Role: Data Analyst
    Objective: Turn data and research into actionable insights.
    AI Need: Discover AI tools, knowledge, and use cases relevant to analytical work.
  • Operations
    Role: Operations Manager
    Objective: Improve efficiency and automate repetitive work.
    AI Need: Find practical AI solutions and examples that can be applied to operational workflows.
  • Investment
    Role: Investment Specialist
    Objective: Support investment analysis and decision-making.
    AI Need: Discover relevant AI capabilities, knowledge, and examples across the organisation.
  • Leadership
    Role: Manager
    Objective: Understand AI opportunities and support adoption within teams.
    AI Need: Access AI initiatives, adoption insights, governance guidance, and relevant organisational knowledge.
  • Technology
    Role: Developer / IT Specialist
    Objective: Explore and implement AI capabilities.
    AI Need: Access technical resources, APIs, development tools, and reusable AI solutions.
  • Country Programs
    Role: Policy Specialist
    Objective: Support country programmes and development priorities through analysis and evidence.
    AI Need: Find relevant research, institutional knowledge, and AI use cases for programme and policy work.

Key User Journeys

The identified needs were translated into role-specific AI adoption journeys, covering discovery, learning, community engagement, responsible use, reuse, innovation, and AI development. Each journey connects a specific employee need with the relevant Portal capabilities and next step, creating multiple pathways through the same product.

  • Discover
    Find and reuse relevant AI capabilities
    • Discover AI initiatives
    • Explore products and use cases
    • Identify relevant entry points
  • Connect
    Build AI knowledge through communities
    • Discover events and communities
    • Connect with AI Champions
    • Share knowledge and experience
  • Learn
    Build practical AI capability
    • Explore learning resources
    • Follow relevant learning paths
    • Apply AI in daily work
  • Align
    Support informed AI adoption across teams
    • Understand AI initiatives
    • Assess adoption and maturity
    • Identify scaling opportunities
  • Govern
    Enable safe and responsible AI use
    • Understand governance requirements
    • Assess AI-related risks
    • Identify mitigation approaches
  • Create
    Move from AI capabilities to new solutions
    • Explore approved APIs
    • Experiment through AI Factory
    • Build and evaluate solutions
  • Scale
    Reuse proven AI solutions across teams
    • Discover existing solutions
    • Access documentation and owners
    • Adapt solutions for new contexts
  • Apply
    Identify AI opportunities in operational work
    • Identify workflow inefficiencies
    • Explore relevant AI solutions
    • Integrate AI into workflows

Product Architecture Diagram

These user needs and journeys informed the product architecture, translating diverse adoption needs into a structured navigation layer for the World Bank’s AI Platform. The Portal brings together core areas, AI Visibility, AI Education, AI Community, AI Governance, and AI Execution, providing connected pathways from discovering and learning about AI to responsible use, experimentation, and AI development.

Beta Release
An initial beta version provided an early product foundation for validating Portal structure, content model, and user experience before further evolution.
PRODUCT EVOLUTION

The Portal is structured around 7 core product capabilities designed to support different stages of AI adoption, from orientation and discovery to learning, responsible use, innovation, and development. This structure translates a complex AI landscape into clear pathways that help employees find relevant capabilities, build confidence, apply AI, and identify opportunities to do more.

1. Overview.
Orientation layer
Context. The growing number of AI products, learning resources, initiatives, and communities created a complex environment for employees to navigate.

Solution. The Portal provides a central orientation layer and continuously curated entry point, surfacing the most relevant and up-to-date AI capabilities, resources, and opportunities in one place.

Value. Reduces the effort required to navigate a constantly evolving AI landscape and helps employees quickly discover what is most relevant right now.
2. Discover.
AI visibility and Reuse Layer
Context. AI products, initiatives, and use cases were distributed across teams and functions, making it difficult for employees to understand what already existed or find relevant solutions.

Solution. Create a consolidated discovery layer that makes existing AI capabilities and their application across the organisation visible.

Value. Enable employees to discover, understand, and reuse existing AI capabilities, reducing duplication and supporting broader adoption.
3. Learn.
AI Literacy Layer
Context. Employees could learn about AI through courses, internal articles, news, blogs, and practical prompts, but these learning opportunities were not connected through a single AI-focused experience.

Solution. Portal curates AI-related content from existing organisational resources and connects it through a dedicated AI learning journey, creating a bridge between learning and hands-on adoption.

Value: Supports employees in developing AI knowledge and moving from awareness toward practical application.
4. Engage Section.
Community and Adoption Layer
Context. As AI adoption grew, employees needed a central space to connect with AI experts, communities, events, and organisational knowledge beyond individual AI products.

Solution. Brings together events, announcements, communities, and training requests, creating multiple ways for employees to engage beyond individual AI products. It supports knowledge sharing, visibility of AI initiatives, connection with AI Champions and communities, and access to learning opportunities tailored to team needs.

Value. Turns AI adoption from an individual activity into a shared organisational capability, helping employees learn from each other, discover relevant initiatives, and spread proven practices across teams.
5. Use Responsibly.
AI Governance Layer
Context. Responsible AI guidance, risk information, and policies were distributed across different organisational resources.

Solution. The Portal brought responsible-use guidance and risk resources into a dedicated part of the AI adoption journey.

Value. Makes responsible AI guidance part of the employee journey rather than a separate governance resource, supporting safer adoption, clearer decision-making, and more consistent application of AI policies across the organisation.
6. Innovate.
Centralized AI Ideas Hub
Context. In a large organisation, AI ideas could emerge across many teams and channels, making them difficult to consolidate and connect with existing initiatives.

Solution. The Portal creates a central space where employees can explore existing ideas, check available AI products, and propose new AI opportunities through a structured submission process.

Value. Creates a pathway from employee ideas to AI opportunity assessment and prioritisation.
7. AI Factory.
AI Development Gateway
Context. Employees and teams exploring AI needed a clearer path from discovering capabilities to experimenting with and building AI solutions.

Solution. The Portal connects users to AI Factory capabilities including the Model Garden, AI Playground, API documentation, and guided development resources.

Value. Provides a clear path from AI discovery and experimentation toward development, making internal AI building capabilities more discoverable to teams considering internal AI initiatives.
Product Evolution
The Portal evolved into a central entry point for navigating the organisation’s growing AI capabilities, connecting discovery, learning, responsible use, collaboration, innovation, and development within one experience. By making AI capabilities and organisational knowledge easier to discover and act on, the product supported a more connected path from AI awareness to practical adoption and new AI opportunities.
IMPACT

The Portal’s business impact was reflected in broader AI adoption, easier access to AI resources, and increased engagement with AI learning and innovation opportunities. Product performance was assessed through Portal usage and user interactions, supported by self-reported user interviews and feedback sessions to understand how employees discovered, learned about, and applied AI in practice.

Enterprise Impact
Primary impact focus: how effectively the Portal is helping employees discover AI, learn, engage, and move toward practical use.

Enterprise Reach
20,000+ employees across 189 countries
Provided a single entry point to AI internal initiatives, learning resources, and expertise across the organization.

AI Adoption ~60%
Around 60% of employees used the Portal during the measurement period, based on Portal usage data.

AI Discovery to Action ~45%
Around 45% of users who discovered an AI capability went on to explore the product, documentation, or related resources, based on tracked interactions.

AI Learning to Practice ~40%
Around 40% of users engaging with practical AI learning reported applying it in practice, based on follow-up user feedback.

AI Innovation Participation ~5%
Created a new channel for employees to contribute AI opportunities, with ~5% of Portal users participating.
MY ROLE
Responsibilities
I worked closely with product management, business stakeholders, engineering, business and data analysts, researchers, and QA across the full product lifecycle, from defining the product context and business problem through product vision and use case discovery to shaping the product design around business and user needs.

Design Direction
Shaped product experience and interaction patterns based on user needs, business goals, AI capabilities, and technical context.

Discovery, Research
Explored user needs, behaviours, and AI adoption journeys through research, user feedback, and iterative validation.

Product Design
Designed product experiences around user needs and journeys, from concept to implementation.

Cross-functional Collaboration
Worked closely with engineering, business and data analysts, researchers, and QA teams to translate design concepts into a coherent and scalable portal experience.

Areas of expertise: Product Thinking · AI Product Discovery · AI Product Design · Enterprise AI Platform · Information Architecture · Technical AI Literacy · AI Adoption · AI Governance


See detailed list of skills
DELIVERY
From Idea to Release
The product evolved through an iterative Agile approach, combining structured planning, cross-functional collaboration, and continuous feedback.

Plan:
Business Context · Product Vision · User Needs · Product Goals · Roadmap · Backlog Priorities

Explore:
AI Use Case Discovery · Requirements · User Research · User Personas · Prototypes

Deliver:
Agile/Scrum · Azure DevOps · User Stories · Figma · Design Handoff · Backlog Refinement · Daily Stand-ups · Sprint Reviews

Learn and Iterate:
Beta Validation · User Feedback · Testing · Analytics · Continuous improvement