Joseph Olassa, CEO, Nuivio Ventures & Ignitho Technologies, with expertise in Data Science, AI & Automation and Product Engineering.
As AI awareness and available technology have evolved, it has become clear that we need to start thinking in terms of ecosystems that service our customers as well as our businesses’ internal processes. In this article, I will explore how ecosystem thinking—thinking in terms of the collection of systems and organizations that execute business processes—can help you build an AI-powered enterprise, and the three ways in which it will likely manifest itself.
Mapping Of Functional Capabilities
The first way ecosystem thinking can manifest itself is through mapping how different functional systems in your company play and connect with each other. Instead of investing in siloed capabilities that are meant to be integrated later when needed, you can review your complete business process and understand how your ecosystem of components will need to come together to meet that process.
For example, at present, business intelligence and dashboarding systems are not typically AI-aware; they are largely based on data exploration and very limited regressions. Similarly, deployment of data lakes is not geared to avoid data duplications down the line, as the data may be needed by different parts of the enterprise to drive AI and business intelligence (BI) systems. Also, more companies are implementing customer data platforms, but machine learning operations (MLOps) and integration of business applications are not typically part of such programs.
This capability-driven deployment, although rapid to implement, can also hinder the effective integration of AI into an enterprise’s business operations. Mapping out the end-to-end process should make it easier to foresee and address such issues, thereby making it easier to operationalize AI.
So, an ecosystems approach can identify complete processes and their interlinkages when it comes to the people using them. This understanding will form the basis of the next step in our approach: design thinking.
Design Thinking
As I outlined in my previous article, design thinking is a must to understand those top-down AI use cases across the enterprise that will be benefitted. I believe that only by involving stakeholders from various departments in an outcomes-driven process will it be possible to create a comprehensive vision for AI integration.
Design thinking takes a view from stakeholders’ perspectives and what they need to be successful. It involves identifying the problems that need to be solved, understanding the context in which the problems exists, and exploring potential solutions that meet the needs of the enterprise.
So design thinking can complement the business process view we developed in the first step. Together, these two steps help to align business needs and ensure that AI is integrated in a way that is useful, usable and relevant for the end user. For example, say you review the need to provide better customer experiences and discover that it is important to personalize, to know when a customer is disengaged or disappointed, and how to identify when they are likely to respond to an offer. With these needs, you can now devise a combination of digital capabilities geared to capture zero-party data and come up with common AI models that must be integrated in real time with all customer channels. The implementation of AI models is not only for digital mediums, though; email and phone calls can be part of the conversation, too.
Technology Integration
The third and inevitable component of ecosystem thinking is how we think about technology. In my view, building an agile AI-powered enterprise must lead to simplification of the data and technology landscape. As we move toward a microservices-based architecture, we do need to think of minimizing data silos and creating a golden source of truth. Successful AI needs quality data.
There are opportunities to select data platforms that are extensible enough without requiring complex integrations. For example, software that offers both business intelligence and AI model hosting instead of requiring these to be separate can help streamline the data workflows and greatly simplify the technology architecture. By doing so, you can potentially reduce the complexity of your business’s data integration and management, which typically takes up 45% to 60% of most analytics investments.
Simplifying the landscape is not an easy task, especially given legacy processes, custom development and the cost of change management overall. There are ongoing advancements in technology that allow for two-way integration behind the scenes, thus making reconciliations much more palatable. So, in my view, incremental simplification should become a part of the enterprise architecture (EA) governance process for any significant initiative.
Today, understanding enterprise data flows and tracing data lineage is a complex endeavor despite the availability of excellent tools. Simplification can help not only integrate these data sources to support better AI adoption but also result in better compliance and reduced risk.
Conclusion
In conclusion, an ecosystem thinking approach can helps us step back so we can unravel the tangled web of processes, systems and stakeholders. This is why I believe that application of ecosystems thinking is crucial toward building an AI-powered enterprise. By adopting the three-pronged approach of functional systems mapping, design thinking and simplification, enterprises can ensure that AI is integrated into the larger business strategy and not just seen as a collection of stand-alone technology components. This can enable enterprises to realize the full potential of this transformative technology.
Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
Read the full article here
