Considerations when establishing an AI Centre of Excellence
If we are considering adopting artificial intelligence, a center of excellence is a good place to start. These centers typically consist of an expert group of technical staff who can advise us on a wide range of AI projects and ensure the highest level of quality. Center of excellence can help us filter out projects that don’t bring immediate value to our organization and focus on projects that do. They should also advise on the impact of AI on business processes and provide technical assistance as needed.
It should lead by example/ Executive Sponsorship
First, the design of an AI Center of Excellence must have executive sponsorship. A CoE can be led by the Chief Information Officer (CIO), Chief Data Officer (CDO), or VP of AI. The CoE should have clear objectives, such as thought leadership, new products, and showcasing AI technology. The design should also include use-cases that can demonstrate the benefits of AI. In short, the design of an AI Center of Excellence should be led by example.
Once in place, a center of excellence can provide governance and standardized AI practices, which will make the adoption of these technologies easier and more scalable. It also provides the framework needed to scale AI processes to meet increasing data volumes, quality requirements, and speed. It can also provide the governance necessary to ensure that the AI practices are used effectively and deliver business value. Ultimately, an AI Center of Excellence can help organizations develop an AI culture that is best suited to their company and its goals.
A CoE requires executive buy-in and a well-qualified, experienced team. If the coE can’t lead by example, it is not likely to be successful.
Create a network of AI champions
To build an AI center of excellence, we must establish a clear vision for success. We must clearly define expectations, essential value proposition, and collaboration. We must also define AI champions, who are the people who will help implement new AI projects. This will help us to identify and attract talent for AI project. A network of AI champions will help make decisions based on shared values.
AI center of excellence must cultivate a network of AI champions. These champions should work across the organization and have a broad range of skills. 45% of companies have appointed a senior executive as AI champions. But these people are not always capable of tackling complex AI assignments. In-house capabilities should focus on statistical and mathematical modeling. If we do not have a data science background, we should consider hiring a consulting firm to help us with this effort.
As an AI champion, we will be able to identify employees who can help with specific projects. AI center of excellence will be able to help find these employees and place them in appropriate roles in the organization. When hiring AI champions, make sure we consider their business expertise. AI experts can assist us with analysis of data and develop prototypes of AI use cases. They can also help prioritize apps within the organization. We need to ensure that they balance the value of the issues with achievable goals. AI center of excellence and company leaders should have a “pipeline” of these projects so that AI projects will be properly tracked.
Develop a talent pool
We should define the roles, set up a clear structure and start looking for new talent. Developing a talent pool is crucial to building the right team and ensuring that we can recruit the right people for the right AI projects. Identifying the right people is the first step in establishing an AI center of excellence. Once we’ve outlined the roles and responsibilities, we’ll need to develop a talent pool to hire the right people.
AI practitioners are in high demand. And with only a handful of top-tier companies, these experts are highly concentrated. While AI experts are in high demand, the shortage of AI specialists has made it challenging for many organizations to attract them. In the U.S., 76 percent of AI PhDs are employed in the private sector or academia, and only 31 percent work in the government. AI PhDs prefer positions that give them a sense of ownership over their research. The AI center of excellence can be a solution to this shortage of skilled workers. Most of the AI research that is being done is open source or proprietary software’s like SAS.
Establish KPIs
To make the most of our AI center of excellence, we must define success. These KPIs should be aligned with the goals of our organization, the principles that will guide future decisions, the essential value proposition, collaboration with internal organizational constructs, and AI champions. Then, we must establish governance and KPIs that will track progress. To achieve our AI goals, company must be committed to making data a central asset for the company. Critically assess our data management processes and determine the best ways to collect, store, annotate, and train it. Make sure to source clean, high-quality data. Moreover, make sure we have a variety of use case libraries that span POCs, standard business unit implementations, and moonshot projects.
Ultimately, the benefits of an AI center of excellence are measurable to our bottom line. But how do we measure its impact? How do we measure ROI?
KPIs should be measurable and aligned with overall business goals. The CoE should use metrics to monitor its success. In addition to these, If we don’t measure these KPIs, CoE will never be truly successful. A CoE must determine how much value it creates. In addition to the economic value of their research outputs, they must determine the ROI of their investments. For example, a large scale CoE may focus on many different areas of research, including sustainable IT, e-government, and machine learning. A smaller scale CoE might focus on a single industrial domain, such as oil and gas exploration.
Develop a process
In order to set up own AI center of excellence, we must identify roles, structure team, and develop a process for hiring new talent. In addition to hiring and training experienced professionals, we must also establish governance to ensurethat the AI center stays focused onn achieving its goals. To achieve this, we must invest in robust data management, validation, and collection to train the AI models and other systems. Once our AI center is up and running, we can start working on the business plan. We’ll need to create KPIs and metrics that will help evaluate progress. Because AI is rapidly changing the business world, expectations can get unrealistic
Deciding on technology to meet short- Medium- and long-term objectives
The technology stack is the cornerstone of a successful AI center of excellence. Select the right technology stack for our business and establish the process for evaluation and adoption of new tools. The right technology stack should be complementary to existing IT infrastructure and align with business goals.
Collaborations are the key to success
The role of the AI center is to orchestrate relationships with external sources of innovation, thereby creating a vibrant ecosystem for AI adoption. To help build the AI ecosystem, we can assign AI analysts to liaisons within IT and end business units. In this way, we can ensure that AI center is part of the organization.
Plan and coordinate research activities: manage external stakeholders
The management team of a CoE should plan and coordinate research activities. This team must define its research target areas and focus on productivity and quality. In addition, the management team should support high ethical standards and pay close attention to responsible research practices. In addition to research activities, CoEs need to plan and coordinate outreach activities. These activities can lead to commercialization and spin-off opportunities. Moreover, the CoEs should be involved in scientific publications and conferences. Their results should be communicated and widely disseminated so that they can benefit society. They should be able to provide technical and educational support to researchers around the world.
In addition to research and development activities, AI CoEs should have external stakeholders. For example, the management team should consider the needs of the host university and affiliated educational institutions. The CoE’s management team should determine which goals should be addressed and how to achieve them. Prioritizing strategic goals is crucial to making the most of the resources available and ensuring that the focus remains on the mission of the CoE.
Data First Approach
Before we begin developing AI center of excellence, determine what types of data we plan to collect, store, and analyze. This will help create an AI-based data strategy. Once we identify data goals, we can select a data platform and tool set that meets those goals. Choose a data management platform that supports all types of data, such as SAS.
The availability of data is a key enabler of AI and a critical value lever for businesses. AI allows organizations to leverage data in order to make smarter decisions and respond to market dynamics faster. Data has become increasingly unstructured and is now generated from more sources. It includes social media and mobile devices, as well as data collected from Internet of Things (IoT). By combining information from various sources, AI can improve the productivity by making better decisions and predicting outcomes. The value levers of AI include intelligent automation, enhanced judgment, and intelligence products and services.
Governance process is mandatory at the end
A good governance framework is essential in AI and a solid AI governance framework helps businesses monitor and evaluate the process. Governance frameworks are vital for in-house AI development efforts and third-party AI tools. Moreover, they help companies monitor their data and identify the effects of degradation or bias on the algorithms
Creating an AI center of excellence means developing a core machine that incorporates all of the learnings and experiences from previous AI initiatives. It also provides a roadmap for how to use AI within our business strategy. Using an AI center of excellence will help team consistently deliver solutions to our customers, creating cost efficiencies and maximizing revenue. In short, the AI center of excellence will give companies an edge in the market.
SAS is working with organizations across the globe to establish Centers of Excellence, and this year we will present cases of such collaboration at GITEX Global in Dubai. If you are there, come by our stand and we will be happy to discuss details.
Happy Learning
DD
