Strategic Management for AI at Scale – Identification 1 of 8

Exploration on Digital Capabilities for AI at Scale continues with Identification, starting with two topics: 1) Introduction to Integration Requirements and Constraints as generic way to identify dependencies between AI Use Cases and Digital Capabilities, and 2) Strategic Management as one of eight digital capability categories. For background, see article Introduction: Digital Capabilities for AI at Scale.

Integration Requirements are about identifying and assessing digital capabilities needed to integrate AI use cases into business operations. Correspondingly, Integration Constraints are about identifying and assessing shortcomings in digital capabilities. In a typical company scenario, it is beneficial to start from constraints assessment to build an overview of digital capabilities’ current state before proceeding to AI use cases and their needs.

In a typical company scenario, it is beneficial to start with Constraints Assessment.

Requirements and constraints can assessed from individual AI use case perspective but also from an aggregate perspective. Article Integration requirements and constraints used Decision Support in Healthcare Services as an example of a single use case to study its integration needs. Correspondingly, Generative AI would be an example of an aggregate need: For AI use cases to be boosted by the latest AI technology, they need to have access to corresponding digital capabilities. In GenAI’s case, to Large Language Models or Multimodal AI Models, and to engineering enabling their integration.

That article discussed use case integration and constraints in somewhat generic terms. This article and the ones coming after discussing Identification will dive much deeper into digital capability areas and their relationship with AI use cases. At later time, Digital Capability Roadmap will be outlined – paving the way to the next main theme in the articles series: Acquisition.

Digital Capability Roadmap will pave the way to Acquisition theme

Finally, previous article on Alignment explored digital capabilities’ general characteristics as strategic objectives, i.e. Scalability, Quality, Speed, Agility, and Innovation – without naming capabilities themselves. Now is the time to dig deeper into digital capability categories, starting with Strategic Management. But before that, let’s lay the foundation by having a quick look at integration requirements and constraints.

Integration Requirements

As discussed earlier in the article series, AI utilization is about embedding AI in all aspects of value creation for higher customer value, better customer experience and enhanced operational efficiency. The term integration requirements is used here as an overarching term to designate various needs emerging from that objective.

In practise, that is about integrating AI use cases into business operations. The term business operations is used here to highlight the fact that besides technical IT related integration, leveraging AI calls for integration within process, organizational and managerial domains.

Overall, there are many different types of integration requirements:

• Direct requirements related to data, AI and software engineering to get the AI use case integration job done.
• Indirect requirements related to platform-like capabilities like cloud computing that support various integration activities indirectly.
• Individual requirements related to single AI use case integration needs.
• Aggregate requirements serving a large set of AI use cases.
• Technical requirements related to technical capabilities like Digitalization, Data Management, AI Engineering and Software Engineering.
• Non-technical requirements connecting to capabilities like Strategic Management, Operating Model and Data Culture.

In practise, the types overlap. The most clear-cut combination would be direct, individual and technical set of requirements. It would look something like this:

• Data Integration requirements: Access to versatile and high-quality structured and unstructured data. Real-time data processing capabilities to cater to use case with immediate response option. Batch processing to identify patterns and long-term trends in data.
• AI Engineering requirements: AI models with full life cycle management from design to implementation and from deployment to maintenance in order to ensure life-long model accuracy.
• Software Engineering requirements: Versatile DevOps methods and practises to enable AI model and data integration with necessary business applications.

However, that is only part of the story. In addition to those “clear-cuts”, typical AI use case integration would also need:

• Digitalization requirements: Natively digital or digitalized products and services, and digitalized business processes to enable AI model integration, including necessary business system updates.
• Cloud computing to support data, AI and software engineering tasks.
• Organizational skills and processes to support engineering and integration, including recruiting and training.
• Change management to update business procesess in order to integrate new AI use cases into existing workflows, and for personnel to adopt new ways of working with the new AI capabilities.

Overall, with the strategic objective of AI at Scale in mind, it is evident that bulk of all the effort and investments needed lies on indirect platform-like capabilities side. That is, to be able to repeat AI use case integration hundreds of times across the whole organization – rather than settling for isolated point-solutions.

Integration Constraints

Integration constraints are about limitations emerging from shortcomings in digital capabilities. The nature of limitations varies from mere nuisance to a total roadblock. Some constraints make AI use case less effective or slows down its development and integration. Another constraint prevents use case deployment all together.

Understanding constraints is essential. Not only because of their negative impact on AI use cases but also to enable systematic constraint elimination.

Constraint types are similar to requirement types. Some would be direct, individual and technical, while some others would be indirect, e.g. limiting platform-like capabilities. And many constraints connect to non-technical capabilities like organizational structure and maturity.

Overall, constraints appear to mirror integration requirements in a one-to-one fashion. That may lead to misguided conclusion that once we are clear on AI use case specific requirements, it’s easy to pinpoint corresponding constraints for elimination. In reality, it is not that simple as constraints’ impact is often systemic rather than point-like. Consequently, Constraints Assessment emerges as essential step of its own rather than mere reaction to use case specific needs.

Constraints Assessment is essential to understand systemic impact and to facilitate systematic elimination

The overall process to deal with integration constraints is as follows:

• Assess constraints within each digital capability area. Document findings carefully.
• Clarify use case specific integration requirements and cross-check with known constraints.
• Prioritize use cases and constraints elimination to facilitate use case integration.
• Eliminate constraints and integrate use cases when the needed digital capabilities become available.

“Constraints elimination” may sound straightforward but in reality it involves complex and often costly effort that calls for holistic take and systematic implementation. In fact, upcoming themes in the article series explore constraints elimination as follows:

• Acquisition – Gaining access to necessary digital capabilities thru outsourcing and in-house build-up.
• Configuration – Organizing and structuring digital capabilities for scalability, quality, speed, agility and innovation while serving AI use case needs.
• Management – Maintaining digital capabilities in accordance with set strategic objectives like scalability while serving current and emerging AI use case needs.

Constraints elimination is more complex than it sounds. Key words: holistic, systematic, disciplined.

Strategic Management

Strategic Management as Digital Capability

The role of strategic management in AI-driven value creation is to set the direction, focus and priorities for the organization. This applies to business strategy and digital strategy alike, including execution planning and implementation.

In this way, strategic management is digital capability just like data management or software engineering would be. Like with more technical capabilities, any shortcomings in strategic management will have negative impact on AI utilization for higher productivity.

AI Use Cases’ perspective

Strategic management appears as a constraint when organization lacks direction, focus and push to stay competitive in the AI-defined competitive landscape. To identify impact on AI use cases, let’s consider use case discovery and integration separately.

Use Case Discovery

AI use case discovery is about identification, assessment and verification of AI as booster for customer value, customer experience and/or operational efficiency. Together, business strategy and digital strategy set targets and provide guidelines for use case discovery. For example, how customer-facing business domains are to look for AI use cases for differentiation, improved pricing power and higher sales margins. Or how business domains overall are to look for AI use cases for higher gross margin thru improvements in operational efficiency. For background, see chapter Strategic objectives for AI use cases in the article Digital Capabilities for AI at Scale – Alignment.

If such strategic target setting and guidance are missing, AI use case discovery is prone to become inefficient and ineffective. The risk of shooting everything that moves without consistency and use case portfolio coherence increases significantly.

To make AI use case Discovery efficient and effective, we need strategic clarity. Clarity emerges from well-functioning strategy process resulting in high-quality business and digital strategy.

Use Case Integration

Beyond merely discovering AI use cases, business domains will also integrate them into business operations. Only then will AI make an impact to value creation. Integration calls for extensive set of digital capabilities, technical and non-technical alike. For wider context, see chapter Strategic objectives for digital capabilities in the article Digital Capabilities for AI at Scale – Alignment.

However, in the absense of clear strategic direction and objectives, it is more than likely that essential capabilities are missing. In other words, they were never acquired in the first place. Later articles in the series will discuss Acquisition in more detail.

To enable AI use case integration, strategic guidance on digital capability acquisition is important.

Constraints Assessment

Constraints assessment on Strategic Management covers four areas with several topics to check and verify:

• Business Strategy – Strategy documentation and its completeness, how AI is connected to strategic targets and priorities, strategy communication, and strategy process.
• Digital Strategy – Guidelines on digital capability needs, acquisition, configuration; How to compete with digital and AI.
• Execution Plan – Documented strategy-driven digital capabilities build-up plan, role of Enterprise Architecture.
• Change Management – What is the implementation capability in terms of organizational learning and change, how organizational inertia is managed.

Constraints Assessment as a Service

Constraints Assessment as a Service covers all digital capability areas from Strategic Management to Data Culture. See detailed Service Description.

AI at Scale Workshop

AI at Scale workshop is a compact one-day event for business and technology executives and managers. Workshop seeks answers to the question: What should we as a business and as an organization do to secure our success in the Age of AI?

Next article in the series:

Digitalization for AI at Scale: Digitalization of products, services, customer solutions, business processes and business systems.