Our organization is a global manufacturing company in CPG Food and we are looking to start our AI journey. Looking for concrete example of what a good AI enterprise strategy looks like. What are the key foundational pillar we must have?
Sort by:
When crafting your AI strategy, it’s essential to align it with your overarching business objectives. Consider whether your growth approach is organic or inorganic and whether your application strategy leans towards purchasing or developing in-house. Reflect on your company’s culture: is it one that embraces innovation and tolerates high risk? Does your workforce exhibit agility and adaptability to change? Evaluate your IT operating model, whether it’s process-based, service-oriented, or capability-driven.
A structured three-step process can guide your journey:
Acquire Knowledge: Understand the AI landscape and its potential impact on your business.
Vendor AI Capabilities: Assess the AI advancements your software providers are making and determine if activating them will deliver value.
Engage Business Leaders: Illustrate the potential of AI to your business leaders and encourage them to pinpoint challenges or opportunities that AI could address.
Consider if your organization already employs AI for data and analytics or if you’re looking to leverage General AI (GAI). Additionally, prioritize investing in Data Governance and Data Quality to ensure a solid foundation for your AI initiatives, provided these areas are already mature within your organization.
I did not believe in a cloud strategy and did not believe in an AI strategy because I don't believe it helps corporate strategy to have separate strategies for general purpose technology. Doing so makes it easy to lose focus on execution and the business metrics. Since AI is a general purpose technology, I believe we are in the realm of tactics and so roadmapping the journey is about breaking down the critical business objectives through analysis into the AI opportunities that can be leveraged. Your strategy execution business plan should be reviewed to assess in the areas of customer product/service design, customer experience, product configuration, flaw detection, fraud detection, or in your business capabilities where it applies.
However, I would not hesitate to look at the SWOT side of your strategy and look at it from a risk perspective. AI is not something that you can market-assess, plan, deploy, control and monitor. AI is embedded everywhere and the digital era has made the corporate boundaries permeable. So, looking, for example, at plagiarism risks, corporate espionage risk, analytical decision making risks, and other areas where AI is already in your door (and welcome) to make contingency plans on how to address the new risks that come along with the benefits of general purpose technologies.
AI should be treated as a resource that supports/delivers other strategies.
Rather than building an AI strategy, you should be identifying how AI can support your broader business outcomes.
In the realm of what a good AI enterprise strategy looks like. The journey commences with defining your problem.
What will be the input, and what’s the desired output? Lay down constraints, acknowledge requirements, and set success metrics. The ACT framework—Alignment, Clarity, Transparency—becomes your compass.
Alignment: Align your vision with goals. Ensure your workflow serves a clear purpose, bridging gaps with state-of-the-art solutions.
Clarity: Bring clarity to your problem’s scope. Define it, understand it, and refine it. Literature review and market analysis illuminate your path.
Transparency: Unveil the intricacies. Share not just results, but insights. Transparency ensures understanding.
With these principles, you can achieve your goal in understanding or defining your AI enterprise strategy.
Please note there is more to unpack on this topic. However, the above will gain you the needed alignment.
Value creation should be the foundational pillar; as you deliver more value to your customers/stakeholders you are simultaneously increasing the overall value of your own organization to customers, suppliers and employees. So how does that translate to technology enablers like AI?
1. First, assess your Data Maturity Level - if most of your data is siloed, that needs to change. The quality of your data will set the stage for successful ML/AI; like garbage in/garbage out, if the quality of your data is poor, the time and cost to do ML/AI is not worth the effort until you have good data quality.
2. People use AI, AI does not use people. How ready is your workforce for AI? What skills gaps or obstacles will you face? What is the goal for shop floor? What will operators do with AI?
3. What automation is already in use. Visual inspection? If you are using VI, move from images to video stream for faster time to value and real time monitoring.
4. If you have an MES is it bought or built? Many MES vendors are adding GenAI but it is not a panacea, and the sector is rife with AI washing meaning what you may be offered may not be all you need. Physics AI +ML+SPC may be the answer.
5. What operational efficiencies need to gained? Reduced downtime, lower cost of poor quality? energy efficiency? See 4, it solves the problems and gets you closer to root cause. GenAI alone will not.
Defining a strategy needs the answers to these questions and many more before you start. AI strategy is like your factories, each is different, and one size never fits all. Word to the wise, flexibility and adaptability are key.