We've proven value, but we're stuck operationalising it
Embedding AI into everyday workflows through people-centred process redesign
This is the hardest part - and you can do it
Moving from "it works in the pilot" to "it's how we work" is where most organisations struggle. McKinsey found only 11% of companies have adopted gen AI at scale. But the organisations that break through share a secret: they focus on redesigning work around AI, not just deploying AI into existing work. We've helped others make this leap, and we can help you too.
of organisations now using AI in at least one function
McKinsey, 2025
of companies have adopted gen AI at scale
McKinsey, 2024
more likely to exceed ROI when scaling at least one strategic bet
Accenture, 2024
"Even when early pilots show clear promise, integration into day-to-day operations is challenging. Senior sponsorship can dramatically accelerate the move from experimentation to scalable success."
Version 1 AI Playbook Guidance
Understanding the Challenge
You've proven the technology works. The barrier now is human systems: processes designed for pre-AI workflows, roles that haven't evolved, and teams unsure how AI fits into their daily work.
IBM's internal transformation provides inspiration. They redesigned workflows into micro-tasks with humans-in-the-loop, achieving 40% cost reduction by putting people at the centre of the design.
IBM achieved 40% cost reduction by putting people at the centre
REACH
Acquiring knowledge and capability
Recognise
Why AI matters now
Engage
Building the will to change
Acquire
How to change
Capability
Skills to implement
Hold
Sustaining change
Your teams need to acquire practical knowledge of how AI fits their work, and the capability to apply it. This means redesigning processes and providing role-specific capability building, not just training courses.
Job Flow Analysis:
Redesigning Work for Human-AI Collaboration
The key to operationalisation is understanding how work actually flows, then redesigning it intentionally. Ask: Which parts could AI do really well? Which parts require uniquely human capabilities? And which naturally go together?
Map Current Workflows
Document how work actually happens today - not the official process, the real one
Identify AI Opportunities
Score tasks for automation potential and human judgment requirements
Define Human-AI Handoffs
Create clear boundaries: where AI acts, where humans validate, where humans decide
Rebuild with Purpose
Design new workflows where AI amplifies human capabilities rather than replacing them
What Success Looks Like
Redesigned Workflows
End-to-end processes rebuilt around AI capabilities—Bain shows this can cut lead times from 60+ days to under 1 day when done well.
Clear Human-AI Roles
Defined roles for AI trainers, validators, orchestrators, collaborators, and decision-makers. Clear boundaries build trust and reduce fear.
Agile Operating Models
Cross-functional teams using iterative delivery. High performers are 3× more likely to have strong performance management infrastructure for AI.
Practical Capability Building
Role-specific learning that builds lasting capability, not superficial familiarity, integrated into daily work, not separate courses.
Practical Actions for Your Teams
- Deconstruct job roles: Break down roles into skill components using SFIA. Which parts could AI do well? Which require uniquely human capabilities?
- Define human-in-the-loop models: Clearly specify where humans validate, decide, and oversee. This builds trust and reduces fear
- Run adoption hackathons: IBM's playbook advocates hackathons with top-down sponsorship to drive hands-on learning and build excitement
- Build psychological safety: KPMG emphasises trust and safety as prerequisites. People need to feel safe experimenting and making mistakes
- Create human-digital pods: Small teams where AI handles routine tasks and humans focus on judgment, creativity, and relationships
