Nearly 40% of businesses remain stuck at the pilot stage. Even when proofs of concept show promise, initiatives often falter due to underestimated barriers in architecture, data, operations and governance. But they don’t have to.
This whitepaper introduces the FOREST framework — six dimensions of AI readiness that help organizations overcome common blockers and scale pilots into production with confidence. You’ll learn why so many AI initiatives fail to move beyond proof of concept, and how to build organizational readiness through:
Build the platforms, processes, and technical foundations needed to support AI at scale.
Align roles, teams, and decision-making structures to enable continuous AI delivery.
Ensure data is accessible, high-quality, and structured to power AI-driven outcomes.
Design AI to augment human work and create value that users want to engage with.
Tie AI initiatives directly to business goals to maximize relevance and impact.
Embed governance, transparency, and ethics from the start to scale AI responsibly.
With this approach, leading organizations worldwide are realizing value across five critical areas:
79.5% increase in basket size for a global fashion brand; 1M users onboarded in 100 days for a new AI-powered service.
75% agent efficiency growth and 20% customer satisfaction improvement at a leading bank.
$15.1M in savings for a major airline; $80K annual savings through process optimization at a global mobility leader.
20% faster customer insights at a global bank; 75% efficiency gains at a pharmaceutical company; 2–5x speed gains through optimized code.
Partnering with the United Nations to promote responsible AI practices worldwide.
Organizations globally have embraced AI and are actively experimenting with it, but the primary challenge is scaling AI proofs of concept (POCs) into production to unlock their full value. Many high-value initiatives can fail to deliver their potential due to various barriers.
The FOREST framework is a model created by Thoughtworks to help organizations understand and overcome the barriers to scale AI, thereby building AI readiness. It breaks down the essential components for AI readiness into six key areas:
• Foundational architecture
• Operating model
• Readiness of data
• Experience for humans + AI
• Strategic alignment
• Trustworthy AI by addressing these areas, organizations can more easily transition their AI POCs into production.
AI development and operations represent significant shifts in how teams work. An effective operating model must evolve alongside technology to prevent misalignment of processes and AI aspirations.
Key barriers include imbalances between specialization and cross-functional collaboration, dependencies between distributed teams and centralized management models.
Building bespoke cross-functional teams and empowering decentralized decision-making can overcome these challenges.
AI should augment and improve human experiences, not replace them. To ensure positive Experiences for Humans + AI, solutions must be designed with a user-centric vision, integrating user feedback and research throughout development and iteration.
Common barriers include humans disliking AI solutions, low perceived user value and difficulty engaging with the AI development process for diverse stakeholders.
Continuous alignment and ensuring solutions evolve with user needs are crucial.
Trusted AI involves establishing robust governance frameworks to ensure the reliability and ethical nature of AI outputs. Without trust, AI solutions cannot be scaled.
Barriers include low visibility into what causes negative AI outcomes, imbalances between automated governance and human input and treating governance and responsible AI as an afterthought.
Implementing evaluation frameworks, safety guardrails, clear responsibilities for human-in-the-loop processes and building governance into POCs from inception are vital steps.
