How to achieve meaningful outcomes from the investment in artificial intelligence capability
Right now organisations are starting to realise that achieving meaningful outcomes from investment in artificial intelligence capability requires a fundamental change in their Operating Model. This means and has always meant people, process and technology. This is comparable to the Agile revolution of the past 20 years.
This article explores the striking emergent parallels between Agile and AI adoption, emphasising the need for a deep technology driven perspective, and the emerging and likely critical role of ‘AI Transformations and AI coaches’ (yes another new job description!) in this transition. It outlines a strategic approach for company leaders and board members to sustainably integrate AI, ensuring both technological excellence and human capability development where we the people actually orchestrate the capability - not the other way around.
Evidence is emerging and suggests that organisations that have successfully implemented Agile methodologies are better positioned to derive value from AI investments, while those that struggled with Agile transformation may face similar challenges in their AI journey. In fact the more Organisational Agility you have, the better your ability to respond to change. And the AI-era might just represent one of the most enduring fundamental changes we will experience in our lifetimes.
"What seems to be emerging is 2025 will be the year many organisations realise that to enable AI will require the same kind of transformation that's required to enable any major new capability."
This observation from a Boston Consulting Group article published in January 2025 highlights the growing disconnect between executive aspirations and organisational reality. While most C-suite executives now prioritise AI initiatives, few organisations are seeing significant value from their investments. This gap exists because many currently approach AI as merely a technological implementation , with a smattering or productivity based use cases, rather than a comprehensive transformation or change management activity where you need to take a system wide view and consider your operating model to truly derive the real value from the immense AI opportunity.
The jury is out on whether this needs to be a true Transformation approach, or a large scale continuous change journey. Kaikaku or Kaizen for this who know their Lean and TPS terminology. Kaikaku means a "radical change" or "breakthrough," contrasting with "Kaizen", meaning "continuous improvement," which are both Japanese concepts related to business and organisational improvement, with Kaizen focusing on incremental changes and Kaikaku on more drastic, transformative shift.
This years 6Degrees AI conference in Brisbane focused heavily on the topic and the general consensus from industry leaders here in Australia is the approach is likely to be more transformative than ongoing change management. Kaikaku over Kaizen.
To achieve the proposed outcomes now being projected by not just consulting firms and technology vendors but also industry luminaries such as Stefan Hajkowicz, Chief Research Consultant at CSIRO, organisations must undertake significant shifts in people, processes, and organisational structures to realise the value from AI potential. Implementing AI tools without addressing the underlying organisational systems will yield limited results at best. The most successful organisations will recognise that any AI Transformation requires a systemic approach that encompasses technology, talent, leadership, and culture.
The global Business Agility Institute just recently published their own report into AI: From Constraints to Capabilities, AI as a Force Multiplier. Their report explores AI's potential as a force multiplier for business outcomes, enabling people and teams to achieve significantly greater business outcomes than they could alone. This impact is highest when AI works alongside people, enhancing decision-making and creativity (augmentation) rather than replacing human effort (automation). This cooperative relationship redefines the role of people as “composers” who provide the inputs, quality control, and creative vision that AI cannot generate on its own.
The findings reveal that even when organisations successfully complete their AI initiatives, they often fall short of achieving their anticipated business benefits. Despite its potential, AI's impact is often undermined by organisational constraints and bureaucracy, such as rigid budgeting cycles, employee skill gaps, lengthy decision-making, and poor data quality. A key data point used is that only 24% of organistions surveyed are currently seeing benefit from their AI investments, with 59% of respondents expecting to see future returns, citing their strategic and cultural readiness.
We have seen this movie before…
The visible parallels between AI Transformation and Agile Transformation are striking and significant. Both represent fundamental shifts in how organisations operate, think, and deliver value. Understanding these parallels helps explain why AI coaches will likely become as crucial as Agile coaches have been in driving organisational change and ensuring that we as a society accept business operating models that still value human centricity - one of the fundamental goals of the original agile movement.
Both Agile and AI Transformations face cultural resistance. Just as Agile challenged command-and-control management styles by empowering teams and embracing uncertainty, AI similarly disrupts established power structures. Traditional managers often resist Agile because it distributes decision-making authority. With AI, we see similar resistance when algorithms begin making or supporting decisions previously reserved for human judgment. This threatens established expertise hierarchies and creates anxiety about role relevance.
Middle management resistance appears particularly similar. During Agile Transformations, middle managers often became the most significant barrier because they perceived the change as diminishing their authority and control though this was often stemming from a place of relevance and status, often ‘encouraged’ by the merit system within organisations itself. The simple overriding and in hindsight wrong narrative was managers are bad, product owners and scrum masters are good. Their resistance typically manifested as "performing" Agile ceremonies while maintaining traditional control mechanisms behind the scenes. With AI adoption, middle managers may similarly feel threatened as AI systems can potentially automate portions of their analytical and decision-making functions, replace humans in the workflow with Agents, resulting in shrinking a fair chunk of their management tasks. Less people, less managers required.
Modernising skills as a major transformation challenge also parallels. Agile required developers and product managers to learn new ways of working – shifting from detailed upfront planning to iterative development and from specialised roles to cross-functional collaboration. AI similarly demands significant re-skilling, requiring people to develop data literacy, prompt engineering capabilities, and understanding of how to effectively collaborate with AI systems. So how we interact with a team of agent workers versus a real human team, or most likely a Hybrid team - a mix of both. In both cases, organisations frequently underestimate the learning curve and necessary investment in training.
Measurement misalignment creates obstacles in both contexts. Organisations often attempted to measure Agile teams using traditional productivity metrics, missing the value of adaptability and customer-centricity. Similarly, AI initiatives are frequently evaluated using conventional ROI measures that fail to capture their transformative potential. This measurement mismatch leads to premature abandonment of promising initiatives when they don't demonstrate immediate efficiency gains through traditional metrics.
The challenge of integrating with legacy systems appears strikingly similar. Agile methodologies often struggled when constrained by rigid enterprise systems and processes designed for waterfall and fixed plan development approaches. AI implementations face comparable challenges integrating with existing technical infrastructure and business processes that weren't designed with machine learning capabilities in mind. At this point in the AI journey, even when you enter the perfect prompt, you can often get the task bungled. Expect that ot change over the next 3-5 years and the task of integrating systems with Agentic workflows will become increasingly simple.
Perhaps most fundamentally, both transformations suffer from the "tool-first" implementation fallacy. Many organisations approached Agile by purchasing software tools and implementing ceremonies without addressing underlying mindsets and behaviours. I heard the same story just recently, where an organisation had rolled out Co-pilot to hundreds of users with little to no training or no outcome - just turned it on. There is this similar pattern emerging with AI, where organisations invest heavily in technical platforms without adequately transforming workflows, decision processes, and organisational structures to leverage these capabilities effectively. Change management will be crucial to future AI success, as it has been for enterprise agile success.
These similarities exist because both transformations ultimately require organisations to fundamentally reimagine how work happens – challenging deeply embedded assumptions about control, expertise, decision-making, and value creation. Organisations that recognise these parallels can potentially accelerate their AI Transformations by applying lessons learned from Agile adoption, particularly the critical importance of addressing cultural and organisational elements alongside technological implementation.
As Dr. Jessica Tayel (Transformation Leadership Institute ) defines it, "Transformation is not merely changing what we do, but fundamentally altering how we think about what we do." This definition applies perfectly to both Agile and AI Transformations, emphasising the cognitive shift required beyond process changes.
In her recent article, Dr Liliya Lund identified this next wave, the Evolution of Business Transformations. Digital to Agile. Agile to to Ai. Defining Ai as the Intelligence layer.
AI-powered automation and decision-making is enhancing customer and employee experience at Increased speed to value through improved efficiency and innovation. The outcome is AI enhances both agile and digital foundations as data is turned into actions and workflows are optimised for maximum efficiency. Digital built the tech; Agile improved how we work; and AI is making us work smarter and more efficiently, while opening the door to limitless innovation.
With the similarities between them now growing in evidence, there are key compelling reasons why organisations need to view AI as a cross cutting enterprise wide transformation effort with very similar characteristics to Agile Transformation, and ot just a adoption of Tooling.
These are just the Use Cases we know today.
BCG research emphasises that successful AI Transformation requires organisations to reimagine their business models, enable their workforces through comprehensive training and support, and meticulously track impact to continuously refine approaches.
The rise of AI agents will further accelerate transformation needs. Two-thirds of senior executives surveyed are now considering autonomous agents as part of their AI strategy. To be fully effective, these agents will require fundamental changes to established workflows, with organisational change management emerging as the critical success factor and perhaps a constraint in many companies.
This transformation is not without risks. As Nobel laureate Daniel Kahneman described in "Thinking Fast and Slow," organisations may become overly dependent on AI for System 1 (intuitive, fast) thinking. This dependency creates vulnerability if organisations outsource too much cognitive processing to AI systems without maintaining human expertise and judgment. The central question becomes: what are the strategic risks of delegating critical cognitive functions to AI for an organisation's long-term viability?
As we have identified striking similarities between Agile and AI adoption, and the why of change, a successful AI Transformation will require critical elements working in harmony. These components address both technical and organisational dimensions, creating a foundation for sustainable change rather than merely implementing new tools.
Successful AI transformation begins with executives who truly understand AI's potential and limitations. Unlike Agile adoption, where leadership could often delegate understanding, AI Transformation requires executives to develop personal AI literacy - many have been on a Data literacy improvement path, and this is will be the next step on that learning journey. There is a need to distinguish between genuine capabilities and marketing hype, and understand the organisational implications of different AI approaches. This literacy enables executives to provide active sponsorship rather than passive approval. Visibly modelling AI adoption, allocate appropriate resources, removing organisational barriers, and consistently communicating the strategic importance of the transformation is critical. When teams encounter resistance or setbacks, executive sponsors need to reinforce the organisation's commitment to the journey. Reinforcing with clarity and communicating aspirations clearly have always been the hallmarks of successful executive teams and leaders - this remains a constant.
Organisations often approach AI as a technology-first initiative, but successful transformations begin with clearly defined business problems and desired outcomes. Each AI initiative should connect directly to measurable business value rather than pursuing technology implementation for its own sake.
A well-structured roadmap typically begins with smaller, high-impact opportunities that demonstrate value and build organisational confidence. This creates momentum for more ambitious initiatives that might require deeper organisational changes. The roadmap should include early wins that build credibility and support for the broader transformation effort. There also needs to be inherent flexibility designed into planning a longer term roadmap. The pace of change is so fast that plan lock in could be fatal. That underlying essence of strategic agility becomes even more important in this endeavour.
Critically this roadmap should bear all the hallmarks of modern governance in that it is visible, dynamic, data driven with implemented modern ERP and Collaboration Tooling giving a strong connection of strategy through to execution and your delivery teams. There’s no longer an excuse to not have a digital tooling infrastructure to aid not just the Transformation, but ongoing delivery effectiveness and assurance.
Data quality, availability, and governance form the foundation of any successful AI initiative. Organisations often underestimate the complexity of preparing data for AI applications, leading to project delays and suboptimal outcomes. A comprehensive data strategy must address data collection and quality assurance mechanisms as well as data storage and accessibility across the organisation. This gets managed by strong data governance policies that balance security with usability. It also demands ongoing data enrichment procedures to enhance AI model performance.
As Martin Chesbrough, our Everest Engineering sage manifesting as an Intern explains, "Organisations that treat data as a strategic asset rather than a byproduct of operations are consistently more successful in their AI initiatives. This requires dedicated resources, executive sponsorship, and cross-functional collaboration."
AI Transformation requires new roles, skills, and capabilities throughout the organisation. Technical teams need expertise in machine learning, data engineering, and AI operations. Business teams need skills in problem identification, use case development, and integration of AI outputs into decision processes.
Beyond hiring specialised talent, organisations must develop broad AI literacy across all functions. This includes understanding AI capabilities, recognising appropriate use cases, and developing realistic expectations about AI performance. The most successful transformations create formal learning pathways for different roles and provide hands-on opportunities to apply new skills.
At Everest Engineering for example, we are already changing how we view roles such as Business Analysts and Product Owners, led by our Director of Product and Design Pilar Esteban. In the article Reimagining Traditional Roles for Product, Design and Delivery, we have embraced a value-driven, skill-based system that prioritises individual value and empowers our teams to deliver exceptional results for our clients. We view product, design, and delivery not as separate functions, but as a spectrum of interconnected skills. Individuals can specialise in a specific area or develop a broader skill set. This shift empowers our people to adapt to new technologies and ways of working and opens the space for Agents to be easily adopted as key enablers to how we are working.
There is also specific emergence of new redefined roles such as The AI Engineer
We have all heard that although AI might take a range of white collar jobs, its also going to create them. The ‘Prompt Engineer’ is one recent example. A Prompt Engineer is a specialist who designs, optimises, and refines prompts to improve the output of AI models like ChatGPT, DALL·E, and other generative AI systems. Their role involves understanding how AI models interpret inputs and crafting prompts that yield accurate, relevant, and high-quality responses. As of March 7, 2025, there are numerous job listings for Prompt Engineers across various platforms including Indeed.com where there are Over 112,000 prompt engineer positions currently advertised.
We now see the growth as well of the AI Engineer - a new breed of technical professionals who bridge traditional software engineering with machine learning expertise and ability to understand the business problem being solved. These specialists combine a mix of strong software development capabilities and a deep understanding of machine learning principles and techniques. Coupled with experience with data preparation and feature engineering and familiarity with cloud platforms and deployment technologies there is a definitive need for “AI Craftspeople”. Adding in the need for knowledge of ethical AI principles and governance requirements and unlike data scientists who may focus primarily on model development, AI Engineers concentrate on integrating AI capabilities into production systems at scale. Their role is essential for moving beyond experimental AI to embedded, production-grade applications and moving at the visionary pace of change being aspired to. Right now on Indeed.com: there are approximately 9,600 AI Engineer positions currently available.
AI adoption fundamentally changes how people work and make decisions, requiring deliberate cultural evolution. Organisations need to develop comfort with probabilistic outputs, establish new relationships between human and machine judgment, and create psychological safety for roles affected by automation.
This cultural evolution must address both resistance to change and unrealistic expectations. Some employees will resist AI out of fear, while others may expect perfect performance that no AI system can deliver. Leadership must actively shape cultural norms that promote appropriate trust, productive skepticism, and collaborative human-AI workflows.
In essence this is in unison with what was required to make Agility succeed in a business.
Inserting AI capabilities into existing processes often creates friction and disappoints users as our over hyped expectations are not met. Successful implementations redesign workflows and decision processes to leverage AI strengths while compensating for limitations. This redesign considers where human judgment remains essential, where AI provides recommendations, and where fully automated decisions are appropriate.
Many organisations make the mistake of automating existing processes without reconsidering their fundamental design. True transformation reimagines workflows from first principles, optimising for the unique partnership between human and artificial intelligence.
The underpinning agile approach of designing for Flow and value streams remains relevant and useful in organisational design focused on AI interactions with individuals. It’s the future that increasingly more process handoffs are between machines and humans. Operating Models will change for AI to succeed.
Organisations need a technical architecture that supports AI implementation at scale, including model development environments, deployment infrastructure, monitoring capabilities, and integration patterns. This architecture should accommodate both organisational solutions and individual team experimentation.
Mature implementations develop clear patterns for integrating AI capabilities with existing systems, managing model lifecycles, monitoring performance, and ensuring compliance with governance requirements. These patterns reduce implementation friction and enable teams to focus on business value rather than technical infrastructure.
Traditional ROI measures often fail to capture AI's transformative potential, especially during early implementation stages. Successful transformations develop measurement frameworks that include leading indicators of adoption, intermediate outcomes, and lagging business impact metrics.
This framework helps organisations maintain commitment during the inevitable challenges of implementation, providing evidence of progress even before full business impact materialises. Effective metrics balance technical performance measures with business and organisational adoption indicators.
Perhaps most importantly, AI Transformations require sophisticated coaching and change management support. Much like the last decade of enabled agile adoption, organisations will need guidance to navigate both technical and organisational challenges, translate strategic vision into practical implementation, and develop sustainable capabilities.
At an Enterprise level, strategic alignment and organisational readiness will be needed to create a supportive environment for experimentation and learning, accelerating the organisation's progress along the adoption curve.
By addressing these key elements in an integrated way, organisations can navigate the complex challenges of AI transformation more effectively than they did with Agile adoption. The organisations that recognise these requirements and invest accordingly will develop sustainable competitive advantages, while those that treat AI as merely a technological implementation will likely repeat the disappointments many experienced with superficial Agile adoption.
In fact if anything the last decade has taught us the way we need to execute this next systemic disruption. Many organisations have adopted agility to some extent and leveraged some value and learnings during the journey that can be harvested for the next transition. The Next Normal.
Just as Agile coaches became essential to successful Agile transformation, AI coaches will play a critical role in guiding organisations through AI adoption. These specialists will help bridge the gap between technical possibilities and business realities, facilitating the cultural and process changes necessary for sustainable AI implementation. Given that a key to the future success of AI in the business will be Operating Model change, AI Coaches will be the linchpin to supporting the purposeful design and transition to the new future way of working, building on the success of agile adoption and using guiding tools and principals such as the Operational Excellent Framework which has evolved from the globally successful Remote Agility Framework. Tools such as these will accelerate new Operating Model designs needed to create future dynamic and adaptive system of work.
Much like the emergence of a hierarchy based agile coach model designed against the varying scales of a business, there is likely to be the emergence of level-setting of AI Transformation Coaches.
Enterprise AI Coaches will work at the strategic level, helping leadership teams develop comprehensive AI Transformation roadmaps, governance frameworks, and measurement systems. They focus on organisational alignment, resource allocation, and long-term capability building.
Team-level AI Coaches operate more tactically, guiding individual teams through AI project implementation, addressing technical and process challenges, and building team capabilities. They help translate enterprise strategy into practical execution while gathering feedback to inform strategic adjustments. Embedded learning and leading by example.
Both roles are essential for comprehensive transformation. Enterprise Coaches ensure strategic alignment while Team Coaches drive actual implementation and capability development.
I can anticipate that investing in AI coaching significantly increases the probability of successful AI adoption by accelerating learning curves through guided experience; and reducing resistance by addressing concerns and misconceptions. Their role will also guide consistent application of best practices across people and teams by facilitating knowledge transfer and capability building via the creation of feedback loops between strategy and implementation.
What we do have in abundance in the workforce today is a large group of people who have been through this process already in the adoption of agility. Its akin to an 80:20 coverage of the skills needed to guide an organisation to an AI Enabled future - with the 20% being the new technology learning required. In other words there is no Greenfield transformation needed. At a system level, we have been here before, we have learned the lessons, and we know how to tackle this new - or next - transformation journey.
AI Transformation touches every aspect of an organisation, requiring governance structures that span traditional silos. Effective governance addresses not only technical standards but also ethical considerations, risk management, and organisational alignment.
Modern governance is a dynamic, adaptive, and value-driven approach to organisational oversight and decision-making. It breaks away from traditional, rigid, and hierarchical models, embracing flexibility, transparency, and collaboration. Essentially it embraces the values of agile as found in the agile manifesto. The adoption of the Govern Agility Stanchions can help to solve for this new paradigm.
Successful AI Transformation requires a holistic approach that addresses technology, people, and processes in equal measure. Organisations that treat AI Transformation as merely a technical initiative will struggle to achieve meaningful returns, while those that approach it as a comprehensive Organisational Change program with humans at the centre will unlock sustainable competitive advantage.
AI Coaches, much like their Agile predecessors, will play a pivotal role in guiding organisations through this complex transition. By bridging technical possibilities with organisational realities, these specialists will help translate strategic vision into practical implementation.
The organisations that thrive in the AI era will be those that build robust engineering foundations while simultaneously developing their human capabilities. This balanced approach ensures that AI augments rather than replaces human judgment, creating a symbiotic relationship between technology and talent.
Collaborate. Deliver. Reflect. Improve. Dr. Alistair Cockburn coined his Heart of Agile approach, to simplify agile, with a radically simpler human centred approach to achieve outstanding outcomes. Nothing could be more important in the AI journey to come.
.png)
