How to Transform a Company into an AI-Powered Organization
Transforming an organization requires strong leadership, expertise, and planning — a strategic roadmap and various initiatives that inject AI into its core processes, services, and products.
But which business functions should get the initial focus — in becoming data-driven and AI-powered? Do companies need AI expertise in-house and how long would it take to develop an internal Data/AI function? Could AI/ML models become a source of differentiation for companies? Or they will be provided as commoditized AI services developed and operated by big tech companies?
Jyotirmay Gadewadikar, Richard Benjamins, Harry Mamangakis, Yasen G. Dimitrov, Aruna Pattam, Jordi Guitart, PhD, Dr. Sunil Kumar Vuppala, Dr. Christian Guttmann, and David Pereira share their insights.
Jyotirmay Gadewadikar
Chief Engineer (Enterprise AI — Systems and Mission Analysis) MITRE • USA
Today it is no longer a question of whether AI-fueled innovation can help drive business value; but how an organization can incorporate change into business operations to drive value, still is. Where do I start, what do I do, how much will it cost, and how long will it take are common questions leaders frequently ask, since a universally applicable AI transformation strategy does not exist. Enterprises are reluctant to begin a significant change journey driven by AI innovation. Enterprises must be aware of the competitors’ use of AI and internal AI innovation and operationalization maturity across business functions. Internally, it is essential for enterprises first to measure AI innovation capabilities and potential and then address capability gaps, if any, to become an AI-driven disruptor.
It is naive to think that enterprises’ efficiency and efficacy can be improved just by adopting AI tools. - Jyotirmay Gadewadikar
It is vital to analyze, formulate, initiate, scale, and optimize an organization’s AI capabilities and subsequent execution through a well-defined value-maximizing strategic framework and operating model that best fits the organization. It is also naive to think that Enterprises’ efficiency and efficacy with using AI can be improved just with the tools and radical approaches without considering sociotechnical complexity and a ‘system of systems’ perspective. Achieving the transformation, though necessary but not sufficient for advanced technology insertion, is a system of systems (SoS) challenge because this SoS is a collection of independent stakeholders and systems external to organizations. These systems and stakeholders are external technology vendors, technology integrators, regulatory agencies, internal organizational stakeholders, and systems such as data silos, human resources, information technology, and leadership. Each constituent system’s goals are separate and may conflict with the purpose of the transformation itself.
For example, technology vendors are driven to satisfy the shareholders and regulatory agencies to improve compliance. This SoS challenge has stakeholders with competing interests, contradictory objectives between participating systems, unsynchronized lifecycles, desperate management, no clear accountability between the separate constituent systems, and no clear escalation routes. The transformation journey can start with assessing and strengthening the internal capabilities described below first[1]:
- Strategy and Leadership. Although cutting-edge technology and talent are needed to drive AI transformation, it is equally if not more important for leadership to align the organization’s culture and strategy to support AI. Alignment of corporate strategy with AI strategy and definition of measurable goals and objectives are necessary to prevent disjointed programs
- Ethics and Governance. Ethics and governance capabilities are needed to define transparency, explainability, appropriate use of data sources, fairness appraisal, and compliance with regulatory and legal requirements. Technical processes for testing the behavior of algorithms through a quality assurance process play a critical role here.
- Data. Data is the seed across all enterprises that allow for AI to scale, and hence, data must be accurate to the extent possible and unbiased to train systems continuously. The more data available, the more the system’s learning can advance. Although data availability and quality may seem like a straightforward concept across organizations, even organizations within the same industry still have different maturity levels.
- AI and Data Science. The maturity of the data drives better outcomes for an enterprise, but it is the methods and the science applied to the data that help draw meaningful insights and make intelligent decisions. This assessment identifies the use of mathematical modeling techniques and the maturity of the methods and models.
- Technology Infrastructure. A solid technical foundation is a critical component for AI transformation and contributes to developing a supporting ecosystem.
- Talent. It is essential to identify how the talent is recruited, developed, and retained and if a skill competency model exists. Enterprises should be capable of adapting and aligning to the new realm that is tying talent operating models to focus on reactive as well proactive recruiting and hiring.
- Organizational Structure. A one-size-fits-all organizational structure will not work because of social and technical variations. Instead, AI adoption and transformation can be driven and governed in several ways — from a centralized center of excellence where C-level executives lead a central group to a decentralized one where the structure is entirely independent of various business units.
- Decisions, Feedback, and Learning. AI-driven transformation can deliver astonishing results only if AI informs decisions. Suppose the relationships between insights and data are nonlinear, complex, and stochastic. In that case, executives may hesitate to make decisions, so it is essential to ensure appropriate decision-making mechanisms exist. Enterprises have only recently started establishing the necessary infrastructure to collect feedback and incorporate learning mechanisms. It is necessary to include multiple measurements to establish causality and identify the best strategies.
Many organizations plan to operate, upgrade, and transform themselves into data and AI-driven organizations. Still, the challenges may be the complexity within that enterprise, both technical and social. Model-based systems engineering approaches can help manage that complexity. Organizations have leveraged a model-based approach to capture, organize, analyze, and synthesize needs and develop a strategy and road map for execution. The intention is to create a repeatable process to optimize and synchronize AI initiatives and investments through the annual planning and budgeting process.
The interdependence of AI initiatives and the required underlying infrastructure in the investment management process is often ignored. To address it, a business process model that connects the to-be business architecture with technical architecture, can help coordinate the evolution of the necessary infrastructure to successfully transform using AI. The overall approach to developing an AI roadmap is based on gathering requirements, getting some validation, making sure they are aligned, and then communicating them across stakeholders — to ensure alignment. The typical process involves gathering information from organizations’ strategic plans and other documents and then engaging with enterprise business lines and service organizations to confirm future strategic business needs and capabilities.
Once the information is captured across the organization, the next step is to synthesize the collected data to develop an organization-wide strategy which would include a logical grouping of earlier gathered requirements into strategic initiatives and to-be-developed capabilities. For example, increasing employee efficiency can be a strategic initiative, including using AI-driven digital assistants and workforce management. Each strategic initiative can be further executed by developing a multi-year investment plan and a road map.
Your AI adoption strategy, must also consider socio-technical complexities and take a ‘system of systems’ approach. - Jyotirmay Gadewadikar
Given that strategic initiatives may cut across multiple businesses lines and operating units, it is important to validate those initiatives across the organization. The next critical step is to align with funding providers and concur on the investment roadmap. Finally, communication with organizational leadership and other stakeholders completes this cycle. This cycle has to be repeated so that the primary inputs from strategic goals are converted into measurable outcomes, business needs are synthesized into strategic initiatives and then executed through a multi-year investment strategy.
Jyotirmay Gadewadikar received the Scientific Leadership award from the US Department of Homeland Security and is an Artificial Intelligence, Decision Science professional engaged in Strategy, Business, People Development, and Thought Leadership. He was previously Chief Product Officer of Deloitte’s Conversational AI Practice and System Design and Management Fellow at MIT.
Richard Benjamins
Chief AI & Data Strategist, Telefónica • Co-founder and VP of OdiseIA • Board Member of CDP Europe • Spain
If you are a conventional company that wants to become a data-driven and AI-powered organization, you will need to embark on a journey: the data & AI journey that forms part of your Digital Transformation. Conventional companies may generate much data as an exhaust of their operation, but are — by design — not ready to collect, store and exploit this data for better decision-making and value creation. The data & AI journey typically consists of different phases including:
- Exploration. In this phase, you will explore a few quick & dirty (business) opportunities to assess whether you want to embark on the journey. This phase usually takes from several months up to a year and requires little investment. Once you are convinced that there is value, you will move to the next phase.
- Transformation. In this phase, you start to organize yourself to become more data-driven. You will perform an analysis to find the most interesting use cases (applications) to start with, considering value as well as feasibility. You need to break silos (technological, departmental, vendors) and collect data across the enterprise into a coordinated platform. You will need to hire a CDO and set up a data team, closely collaborating with IT and HR. You will need to work on a ‘single version of the truth’ so that the whole enterprise understands data in the same way without ambiguity. This phase will require significant investment and may take between three and five years. Be prepared to have some patience as an organization.
- Data-driven. In this phase, you will start to enjoy the results of your endurance in the first years. You will be able to use data in a consistent way to inform the big decisions you have to make related to your core business. Moreover, new data-driven products and services will see the light. Depending on your sector, you might be able to externally monetize your data and insights to other sectors in a B2B business model.
- AI-empowered. In this phase, you will use Machine Learning and other AI technologies to scale the value of data throughout your organization. Given the massive scale of use, mastering data privacy and AI ethics become essential to create and maintain trust with your stakeholders.
In each of those phases, you will have to make many decisions that will determine how fast or slow you will progress on your journey. Those decisions are different in nature and relate to various aspects of the organization such as business & finance, technology, people, and responsibility.
- On the organisational side, you have to think about where to place the Chief Data Officer, how to measure data maturity and what will the relation be between the data and the IT department.
- On the business and finance side, you need to make decisions about how to select the best use cases, how to measure economic impact, and how to finance the whole data journey, which can take years. There is no single right answer for the question of what business function to start with. In the beginning, it is important to choose a business function that matters for core business but that is also feasible from a practical perspective. Otherwise, it will take too long to provide the first results. Having said this, many organisations start with marketing.
- On the technology side, important decisions include whether you want to work in the cloud or on-premise, whether you need a unified data model, and you need to define a data collection strategy including planning and budgeting.
- On the people side, it is important to create a team with the right skills and expertise. Many organisations choose a mix between hiring expert personnel and training existing personnel. Sometimes they outsource the first initiatives to a third party to kickstart the activity with the objective to later internalise the knowledge. Other decisions related to people include how to democratise all the data initiatives, how to win over skeptics, and how to make people enthusiastic about data through appropriate communication.
- Finally, on the responsibility side, you need to understand the social and ethical challenges of AI and Big Data; to define AI principles and implement them in your organisation; and work out how you can use data as a force for good, to improve society and fight its challenges.
Dr. Richard Benjamins is among the 100 most influential people in data-driven business (DataIQ 100, 2018). He advises the European Commission and Parliament as well as companies and start-ups. He has a passion for AI for Good and authored three books on AI.
Harry Mamangakis
Chief Technology & Operations Officer • Voiceweb • Greece
Transforming into a data-driven, AI-first organization, will always be an ongoing journey since it is part of the overall Digital Transformation Initiative which calls for continuous improvement and adaptation (being agile as an organization). In a typical transformation scenario, which is well documented in the book ‘A Data Driven Company’ by Richard Benjamins, we can consider 4 general states in this journey:
1. Exploration
2. Standardization & Transformation
3. Data democratization
4. AI-first
What happens in each state, depends on each organization and there is no golden rule for all. This means, that each organization will need to have the tools and metrics to measure its fluency in becoming a data-driven and AI-first company. Measuring its fluency helps understanding where the organization is and what needs to happen next.
Transforming into a data-driven, AI-first organization, is an ongoing journey. - Harry Mamangakis
In the ‘Exploration’ state the organization usually attempts to run ‘Proof of Concept’ (PoC) for specific business (use) cases. Usual candidates for such PoCs are marketing campaigns with the goal to improve their effectiveness or use cases like predicting and reducing churn (where applicable) or increasing average order value by cross-selling related/recommended products. From an organizational perspective, an owner is assigned, sometimes also called a champion. In general, this is a bottom-up approach, which means that those involved could be some managers that favor experimentation and data enthusiasts — typically data engineers and/or scientists. There is no formal organizational structure in this state — as this is seen as a specific initiative. This process can repeat over and over with the same or different teams; however, the organization is still considered to be in the “Explorer” state as this happens.
An organization has reached the ‘Standardization & Transformation’ state when there has been an executive decision that data is to be treated as a strategic asset, and that customer value creation must use data in a systematic way. Having reached this point, the organization already has a backlog of use cases to be addressed, and this backlog has been prioritized from the top and communicated throughout the entire organization. Two major initiatives take place during this phase: (a) Data Standardization and (b) Organizational Transformation. To deliver on these initiatives, the following must be there:
- Data Sourcing Strategy. This strategy is about what data to use, and where to find it but it also aims to address organizational difficulties where certain functions consider data “their property”. Thus, the Data Sourcing Strategy also includes how data is to be shared across the organization.
- Discrete Budget. In this state, each department needs to explicitly state their data, analytical or BI requirements, and this is to be approved in the annual budgeting process.
- Formation of a ‘Data Team’. In this state, the organization has realized that a special Data Team is required, with a Chief Data Officer heading it. The positioning of this function within the organization is a whole other discussion, but in general, if we look at successful initiatives, we will see that the CDO is usually placed in organizational structures that are horizontal and apply to the entire business, such as IT or the Digital Transformation Team or under the COO. In some cases, the CDO can also be under the CEO but again, this is a topic for another discussion. The Data Team will consist of data engineers and data scientists and will have worked with IT on the technological choices required to set up the tools to perform the required work.
The next state is ‘Data democratization’. An organization has reached this state, when the use of data is included in the normal decision-making process, in addition to intuition, experience, or expertise. In this state, the Data Team has matured and has delivered the ‘Data Architecture’ for the organization. This includes:
- The inventory of data sources which is always kept up to date.
- A data dictionary to ensure a common ‘language’ and understanding across the organization. In addition, the data dictionary ensures data is traceable to its source, is granular enough and there is only one version of it.
- Processes for data ownership and stewardship to ensure data is of the required quality, always up to date, and available to all those required.
In this state, the organization has already acknowledged that a lot of processes need to be adapted. Hiring needs to change to include skills such as ‘use of insights for decision-making’. Training programs need to be introduced to allow existing staff to acquire the new skills required. Training is very important because, in this state, employees are empowered through ‘self-service’ (processes, tools, knowledge of data, etc) to use data and insights in their daily tasks. Finally, for the Data Team, this state flags the transition from the focus on data engineers to data scientists, meaning that the number of data scientists at the end of this state, must surpass that of the data engineers.
‘Data democratization’ has been achieved, when data is used in regular decision-making. - Harry Mamangakis
The final state is ‘AI first’. In this state, the organization uses Machine Learning (ML) and other AI technologies to create value. This is the state where the organization will also reflect back and decide if adjustments are to be made to the data strategy, organizational structure, and so on. Having reached a level of data maturity, the organization will not hesitate to use ML for direct interactions with customers, for example for personalization, product recommendations, or chatbots (NLP); such decisions will be Business As Usual (BAU). The Data Team will be enhanced by AI engineers (e.g., ML engineers, ML researchers) and transformed into a Data & AI Team. Organizational structure is critical in this state. AI talent and resources should not be working in a silo but should be part of agile end-to-end teams, besides the Product Owner and the other roles required to deliver a product or service. This will enable them to contribute their expertise in real-world situations and not operate in a vacuum. This agile structure is what makes an organization an AI-first organization.
So how does an organization measure its data fluency in order to assess where it is positioned in the data transformation journey and what needs to be done next? There are four dimensions required to be measured, to assess the organization’s fluency:
1. The Technology dimension. What technology choices are made for tools and platforms, how they are used, and by whom. Budgets will need to exist for all these tools, platforms, and so on.
2. The Data Management & Governance dimension. This includes data protection, legal compliance, data security (encryption, anonymization, access control, etc). Data management also includes the functions of Data Architecture as previously explained.
3. The Organizational Dimension. Changes the organization is making to adopt the use of data and AI.
4. The ‘Business Dimension’. This measures the adoption of using data and AI throughout the organization to help in decision-making and optimization of processes but also includes efforts (such as R&D efforts) to use insights for new business opportunities.
Measuring each of the above dimensions will help the organization understand where it needs to adapt.[2][3][4]
Harry Mamangakis is a Technology Executive for over two decades, balancing between fluency in technological breakthroughs and having a business mindset. He has led and participated in several transformation engagements for leading brands in industries such as Telcos and Retail.
Yasen G. Dimitrov
Co-founder & Chief Analytics Officer — Intelligence Node • UAE
As businesses gear toward a post-pandemic world, AI adoption across industries has rocketed — with the pandemic pointing out glaring inefficiencies in conventional operations and the evolving economy demanding more automation, efficiency, and data-driven decision making. With the accelerating number of AI use cases (from simple image recognition to AlphaFold[5]), a lot of sophisticated AI solutions have become available in the market and continue to evolve. While a lot of companies are successfully leveraging AI at an organizational level and have seen tangible improvements across bottom-line revenue, productivity, and cost savings, the majority have failed at implementing AI at scale. These cases tell us that while transitioning to an AI-led organization seems like the obvious next step, it comes with its own challenges and needs a strategic roadmap to see any success.
Conventional organizations need to first analyze where they need AI applications and if they do need them in the first place. Companies need to understand in what capacity to leverage AI and what the short and long-term goals for AI implementation need to be. The first step to moving toward a data-driven, AI-powered organization is to take stock of the internal data capabilities and talent and start with building a ‘Data and Analytics’ team.
Artificial Intelligence and Big Data bring a paradigm shift in how we do business. - Yasen Dimitrov
To do so, companies need to create an ownership function, starting with the Chief Analytics/Data Officer, who will report to the CEO (this is very important because otherwise activities can be steamrolled either by the CFO or COO). As a second step, a small team needs to be hired, with the following skillsets:
- Senior Analyst (to collect and analyze company use cases and create internal POCs)
- Data Engineer (to put all data sources in one place — internal, eg. Point of Sale data, competitive data, etc)
- Data Scientist (to start using external services like Google Cloud etc. and build some models to complete the POCs)
Depending on the POC outcomes and the expected ROI, the company may hire more of the above, eg., if more custom models are required, hire more data scientists.
However, finding the right talent and training them can be expensive, let alone building an internal AI infrastructure from scratch. Hence, if you are not a ‘heavy data-handling company’, using external APIs to achieve your goals is key. Partnering with vendors that specialize in AI solutions that you are seeking can be faster (through API access), less risky, and more economic. But if the cost of the external APIs is too high and there is no ROI, the probability for you to succeed by internalizing the project is very low as you won’t have the right talent. It is therefore important to strike a balance between partnering with the right Artificial Intelligence vendors that have a rapid ROI potential and building your own internal capabilities.
The next step is to identify which business area or function to target first. Rather than trying to introduce AI across organizational processes in one go, it is much more judicious to analyze the simple challenges first and identify the low-hanging fruits where AI can be easily integrated and create significant value.
For example, an AI-driven pricing solution will always start with a simple regression model tested on a small set of SKUs. Once this process is complete, review the ROI and analyze the business impact it can create at scale. It is important to understand that it is an iterative process that requires frequent reviews.
It is important to bear in mind that transitioning to an AI-driven organization takes time. Building AI applications is a continuous process and works better if you follow a test-and-learn approach to identify and resolve problems early on. Setting up and implementing AI processes cannot happen overnight. Finding the right talent to build a focused, in-house team with domain knowledge and then forging relationships with vendors to partner with, can easily take up to a year or even more in some cases.
AI and big data analytics are bringing a paradigm shift in how we do business — helping improve revenue, productivity, market share, and processes across departments and making organizations future-ready. To take full advantage of data analytics and AI, companies need to incorporate these technologies into their vision and core business processes. Companies need to be adaptable, and flexible; they need to realign their culture and conventional processes to make place for AI-driven decision-making.
Yasen Dimitrov is the Co-founder and CAO of Intelligence Node. Yasen has a proven track record of building BI and Predictive Analytics solutions across various industries and has experience in converting the ‘value’ of data into real, tangible business opportunities. In his current role, he manages operations, analytics architecture, and category expansion; including building and maintaining the largest database (1.2 billion unique retail products) in the industry.
Aruna Pattam
Head of AI & Data Science — Asia Pacific, Middle East, and Japan — A Global Technology Company • Australia
Artificial Intelligence can help a business adopt a more data-driven approach to decision-making. By harnessing the power of big data and analytics, businesses can gain a better understanding of customer behaviours and preferences, as well as market trends. This, in turn, can help them make more informed strategic decisions about where to allocate their resources and how to best serve their customers. AI can also help a company automate many tasks and processes, thereby freeing up employees’ time for more value-added activities. This would not only improve efficiency and reduce costs, but it would also enable employees to focus on more strategic tasks that require human interaction and judgement.
So, how can a company go about making the transition toward an AI-powered organization? There is no one-size-fits-all answer to this question, as the nature of this transition varies depending on the specific needs and circumstances of each business. However, there are a few key steps that all companies can take to get started:
- First, it’s important to assess your current business environment and identify areas where AI could be most beneficial.
- Second, once you have a clear idea of the opportunities that AI brings, you can develop a plan for how to integrate it into your operations.
- Third, the business would need to create a data-driven infrastructure, possibly on top of a data lake, a data warehouse, and so on.
- Fourth, the business would need to invest in AI technologies, and find ways to use these tools to improve its operations.
- Fifth, the business should set up a dedicated team to manage and oversee AI initiatives, or develop partnerships with external service providers.
- Sixth, the business would need to adopt a data-driven culture. This means that all employees would need to be data-literate and understand how to use data to make better decisions.
- Finally, it’s important to remember that a successful AI transformation requires a commitment from all levels of the organization. Employees across the hierarchy need to be on board with the new strategy and be willing to work together to make it a success.
If a company can successfully make these changes, it will be able to harness the power of big data and become a data-driven, AI-powered organization.
But how could a company find the right talent and form a powerful Data & AI team? One option is to hire external service providers who have the necessary skills and experience. This can be a cost-effective way to get up and running quickly, as many providers offer a range of services that cover everything from data management to AI development. Another option is to develop an in-house team of AI experts. This can be a more expensive option, but it gives businesses more control over the development and implementation of AI initiatives. A company could also choose a hybrid approach, which involves hiring external service providers for specific tasks while developing an in-house team for other tasks.
No matter which approaches a company chooses, it’s important to make sure that the team has the necessary skills and experience to successfully implement AI initiatives. This includes skills spanning Data Science, Machine Learning, and Artificial Intelligence. When forming a team it is important to ensure that there is a diversity of backgrounds and skillsets, so that a variety of viewpoints can be considered when making decisions about AI.
The time needed to set up a Data and AI function varies depending on the size and complexity of the company. However, a good rule of thumb is that it will take a minimum of 12 months to get a functioning team in place. This includes hiring and training the right people and getting them up to speed on the company’s data infrastructure and AI technologies. It’s important to note that a Data and AI function will require ongoing development and support, so it’s important to set aside adequate resources for this.
AI has the potential to become a source of differentiation for modern products and services. This is because big tech companies are not well-positioned to provide custom AI services to businesses — they are more focused on developing general-purpose AI technologies that can be used by a wide range of businesses. However, businesses need specialised AI services that are suitable for their specific domain and this is where a company’s in-house team of AI experts can come in handy, as they can develop custom AI solutions, tailored to the needs of the business.
The benefits of a data-driven, AI-powered organization are clear: increased efficiency, better decision-making, and a competitive edge in the marketplace. However, making these changes is not easy, and it requires a commitment from all levels of the organization. A successful AI transformation is a continuous journey. This means that a company needs to invest in AI and get prepared over the long term, and not just in the early stages of the transformation.
Aruna Pattam is a Global AI Thought Leader with 22+ years of experience in data analytics, AI, and ML. Also, a speaker, mentor, blogger, vlogger, and women advocate with the goal of educating and providing awareness on AI to the business and the wider community. The views expressed are personal.
Jordi Guitart, PhD
CEO — Science4Tech Solutions • USA
Digital transformation brings companies seamless access to an increasing amount of data that is readily available in real-time — where real-time here means as soon as the data is generated is eventually validated. Only companies that have completed this digital transformation are able to tackle the next step with a minimum guarantee of success, that is, injecting AI into its core processes, services, and products.
But AI is a relatively new technology — well, in fact, it can be seen as a sleeping beauty since the late ’50s of the twentieth century, until GPU’s computing power was capable to shake and awake the AI technology — and it demands an entirely new skill set that it is typically not available in-house. Roles like Data Scientists, Machine Learning Engineers, and Data Engineers, are not so abundant out there, or at least their number is not growing as fast as industries need them.
Additionally, companies need new corporate and senior management roles to encompass the necessary knowledge. Of course, companies can rely on external AI know-how. Consulting companies are out there to provide the basic understanding of what is AI and tell you the dos and don’ts. Externalizing efforts is a well-known way to mitigate risks for those companies that believe AI is not — and could not — become key to their business.
The first thing companies must consider is hiring an AI leader to build a core AI team. - Jordi Guitart, PhD
For those companies that are successfully evolving through their digital journey though, my recommendation would be to capitalize on AI and the huge volume of data they are digitally gathering, as these will be the two new pillars to sustain companies’ long-term value generation. No matter the size of the company — particularly thinking of large non-tech companies — I personally advise against going big in AI from day 1 to any company aiming at getting AI as its main source of differentiation. It’s all about knowledge, not size! So, building a small AI team from scratch and scaling it up upon success will pay off in the long run. Hence, the first thing companies must consider is the hiring of an exceptional AI leader so to let him/her build his/her AI team. No matter the titles the AI team leaders wear, they are the only ones able to attract, motivate and retain AI talent around them. Of course, developing an in-house Data/AI function to create a source of competitive advantage takes time.
Despite that growing an AI team following a slow-paced strategy may sound counterintuitive compared to the overwhelming volume of data at reach, experience tells that AI teams addressing many problem statements in parallel is totally inefficient. AI is the result of forward and backward thinking that is intrinsic to any learning process leading to superior knowledge. So, given one problem statement to solve at a time, I strongly recommend letting the AI team dive deep into the data and only ascend while consolidating the learnings, like scuba divers respect decompression stops. Moreover, the development of good performing AI/ML models cannot easily accommodate software coding sprints so agile methodologies are not deemed appropriate for AI teams to adopt.
Likely, the above explains why big tech companies are acquiring small but highly specialized AI companies to faster merge their teams while incorporating new functional knowledge and thus accelerating AI. The impact of high-growth strategies that big tech companies are experiencing is that they are falling fast into commoditized AI products and services that do not differentiate from each other. The lack of technology challenges could explain why the retention of talent in AI is becoming increasingly difficult and only the lure of large salaries seems to be the alternative that Big Tech companies have, thus creating an excessive turnover and wage inflation.
Jordi Guitart, PhD has been recently appointed CEO of the Barcelona-based healthtech startup Science4Tech Solutions, coming from Aizon where he served as VP of AI. He is concurrently Adjunct Professor of Strategic Management at ESIC Business & Marketing School.
Dr. Sunil Kumar Vuppala
Director, Data Science — Ericsson R&D • India
Customer needs should be at the forefront of everything a company does and, as a result, detecting changes in consumer behaviour is critical to a company’s success. Digital transformation helps companies address this by adapting the way business gets done and by creating entirely new classes of businesses. This change requires the transformation of both internal (employee-facing, operational) and external (customer-facing) functions of the organization. Such transformation can improve existing capabilities, for example, using business insights, building relationships, or improving business processes; it can also lead to the creation of new products, services, and business models. Becoming data-driven and adopting AI is also an essential part of this transformation. Overall, it is the 3Cs — Computing, Content, and Communication — that play a significant role in the transformation journey.
As part of the initiative of transforming a ‘conventional’ company into a data-driven, AI organization, business leaders must have a solid AI-focused data strategy and roadmap. As part of the roadmap, leaders need to identify the problems, i.e. the pain points of the organization, and also the opportunities, i.e. where there is space for better decisions by using data and identifying actionable patterns. The process of shaping this AI-focused strategy, includes the formulation of an enterprise data architecture, the definition of a roadmap after understanding the current state, and the use of AI playbooks, with an initial focus on proof of concepts that attempt to address the most pressing business problems. It is also critical to develop a strategic data acquisition policy and introduce the right technologies, e.g. a data lake and data management tools, as well as data governance in line with the data and privacy standards.
After demonstrating the value of AI to the senior management by identifying and executing the appropriate set of early use cases, leaders should focus on scaling up the experiments and expanding the team in accordance with the organization’s data and AI strategy. However, building AI expertise and the right team is a long process. Thus, companies must identify their core competencies and build their team surrounding those key capabilities. But, at the same time, they should leverage their partners, who can offer expertise on a variety of AI platforms and technologies to accelerate the transformation process.
AI adoption and democratization can be accelerated by using platforms and AI as a Service (AIaaS) from big tech companies such as Microsoft, Google, Meta, and Amazon. Companies can leverage such commoditized AI services to drive the differentiation of their products and services. However, the transformation of the company itself depends on its core business and how the technological change affects its operations and customers.
A ‘big bang’ approach for such transformational initiatives is known to have failed in delivering the expectations on various occasions. Hence I would recommend that the company follows a step-by-step process, always emphasizing the demonstration of AI’s value and potential to the senior leadership of the organization.[6]
Sunil Kumar is a global thought leader in AI, IoT, and Analytics; Director — Data science; Top 10 Data Scientist in India; ACM Distinguished Speaker; Inventor of 40+ patents, Visiting Professor, Fellow of IETE, IEI; Technical Role Model and IEEE Engineering Manager of the Year awards winner.
Dr. Christian Guttmann
Executive Director, Senior Researcher — Nordic Artificial Intelligence Institute, Karolinska Institute • Sweden
Every conventional company has products and services offered to industries ranging as wide as manufacturing, financial services, health care, and transport. These industries change constantly and that change is often driven by customers that expect better quality at a reasonable cost. Companies that adapt to changing conditions and meet customer expectations will continue to grow their business and stay relevant in a competitive business landscape.
Create an open space for transformational possibilities for your company’s future. - Dr. Christian Guttmann
One of the most important success factors of the AI transformation journey is the role of the board and executive leadership, as it is their primary responsibility to set the direction and respond adequately to changing market forces. For many conventional companies, Artificial Intelligence (AI) proves to be a major force as it can transform a market fast. Customers quickly change their demands as emerging AI-driven products and services provide superior quality and performance. The most powerful decision-makers of a company, namely the board and executive leadership, need to understand the potential of AI — as otherwise the company may face fierce headwinds, and could be outcompeted quickly.
Machine Learning (ML) is one of many AI technologies that has transitioned from research labs into business use. The application of ML has already generated multi-billion USD in consumer markets (e.g., online marketing and content recommendations) and ML is well on its way to transforming traditional ‘non-AI’ and ‘non-data driven’ industries. By addressing the following three topics, the board and executive leadership may benefit in making the right decisions.
- How do AI and data-driven technology change a conventional company’s market position? This is an important question as a conventional company could be directly and immediately affected by competitors that use AI more effectively than they do. The size of this threat needs to be assessed while taking into consideration what is hype and what is real. If a market changes over a longer period then this is often due to long cycles of product introduction in a particular industry (e.g. the pharmaceutical and car industry). In these long-cycle industries, regulations and new frameworks will take time to put into action, but when they are, it often opens a flood gate of new products, and it is important to be ready when that happens.
- How can AI and data-driven technologies add value to its products, services, and operations? There are at least three AI value generators in a company: AI can improve existing products and services, AI can create a new portfolio of products and services, or AI can make operations more efficient and effective. Operational improvement is often the initial way for a conventional company to create value as the operations do not affect the main value creation of their business — it is, however, often a more conservative path. In operational practice, AI can improve processes, such as logistics and supply chain, sales, and IT processes much of which resides in ERP and CRM systems. The company’s services and products will then be created more effectively and efficiently, hence keeping the company competitive. In this context, it is worth mentioning the well-known ‘productivity paradox’, where a new technology does not seem to have a measurable ROI, that has an impact on the bottom or top line of the business. However, the introduction of AI keeps a company competitive, meaning that they can continue to build and sell their products and services under competitive operations. Successes might then result in spillovers into the first and second AI value generators mentioned above.
- How to assess the risk and cost of introducing AI technologies that bring wide implications on products, services, and operations? It is critical for the company to estimate the risk and cost that comes with AI transformation, as every company has a unique risk profile and will have an acceptable level of risk and cost. A low-risk approach often results in smaller market shares, less growth, and less future revenue, and that path is often chosen if AI technology is not well understood. Deep knowledge of AI in the board and executive leadership is valuable in predicting market developments and identifying what changes need to take place in the company.
These are tough questions, but as history has shown, companies that understand how new technology changes a market and adapt sufficiently quickly will prevail over those that do not. This has been the case with the steam engine, transistor, internet, and electricity. The need to understand and adapt will be no different for a business in the age of AI.
Each board and executive leadership will respond to the above questions differently, and influence a company’s path in their own way. However, there are three considerations that can help to address the above.
- Ensure that there is deep subject expertise for an adequate understanding of AI in the context of your company’s future success. Executive leaders and board members should interact with the smartest and most suitable AI leaders in the company as much as possible to gain a comprehensive understanding of what AI can do for the business. Such a leader has a PhD in AI and a strong business acumen. This will help to identify which business direction to take and what is technologically feasible. In short, hire the best AI talent that you can get your hands on.
- Ensure full awareness of the strength and weaknesses of the company’s existing product and service portfolio. Conventional business leadership is often caught up in group think where everyone agrees that “our business has been good so far, so let’s do more of the same a little bit better, and we will be fine”. This is harmful to a company’s adaption of new market conditions. Hence, create an open space for transformational possibilities for your company’s future.
- Create meaning for key stakeholders in times of change. Initial AI projects or products should aim to be meaningful, rather than impactful, for as many stakeholders as possible. ‘Meaning’ here suggests that the application of AI in the business makes integral sense to key stakeholders that are often resistant to organizational change. These stakeholders can be middle management, employees, and shareholders who are sometimes not convinced of the value that new technology can generate (particularly when their own department or job is perceived to be under threat). For those stakeholders, the outcome of an AI initiative needs to meet a wide range of minimal acceptable — and often intangible — criteria. In practice, an initial AI product or project does not have to be the one that creates the most value, but the one that convinces important stakeholders about AI technology.
Artificial Intelligence is possibly one of the most transformational technologies that humanity will encounter. Transforming a conventional company requires the board and leadership to make a deep evaluation of the company’s business and apply the right expertise to identify how AI can drive business value and future success.
Dr. Christian Guttmann is an Artificial Intelligence scientist and global AI executive with a strong track record in creating high-impact business and research with Artificial Intelligence, Machine Learning, and Data Science. Christian was recently named a Top 100 AI global leader by Deep Knowledge Ventures, based on my achievements in science, technology, and business.
David Pereira
Head of Data & Intelligence — NTT DATA EMEAL • Spain
This question goes beyond the ‘typical roadmap’ to whether the company has the strategy to achieve this transformation to an AI-powered organization. Such a strategy must be nurtured by advanced technology and data capabilities as the means of achieving this transformation. This strategy is paramount, it is the backbone of the organization for [a] generating new revenues, through new value-added services, new business models, and improved customer experiences, and [b] increasing efficiencies, through automating processes and scaling services by leveraging data and AI.
Organizations must establish training programs to develop the internal Data & AI talent. - David Pereira
As the Head of Data & Intelligence in NTT DATA EMEAL, I believe that the best way to support a company in designing such a strategy is through the implementation of an end-to-end framework for Data & Intelligence. This framework should include the methodologies, organizational principles, and technologies that, when combined, can bring to a company all the needed capabilities, at scale. We use this framework to analyze the organization from a strategic, organizational, and operational point of view — we seek to inject Data, use AI and bring innovation into all key processes, and align the efforts among the different areas.
To become a data and AI-powered organization, the end-to-end transformation must permeate all organizational layers from the C-level and senior managers throughout the hierarchy. Each level has a critical role to play in this transformation journey. For instance, the CEO is responsible for defining the strategic lines and setting the corporate objectives regarding the use of data and AI. The HR team communicates and evangelizes the new ‘data mission’ as designed from the C-level, and promotes a culture that emphasizes the use of data. The latter can be achieved by introducing training and literacy programs and certifications; or by encouraging access to the open ecosystem to keep up to date with the latest developments and new trends related to data and AI.
In parallel, business functions play a key role, as they are expected to create value through innovative use cases that leverage Data and AI and bring together operational, industry, and technology expertise. To implement solutions that support these use cases, business functions need to collaborate with technology teams — to build and scale data-driven solutions. Ideally, the first use cases to implement should have low complexity but a high return on investment. This helps obtain support from stakeholders and demonstrate the value of the technology and also helps drive the demand and readiness for more ambitious data-driven use cases across the organization.
Besides fostering multidisciplinary teams and promoting synergies between business areas, an AI-driven organization must establish training programs that help develop further the internal Data & AI talent. These programs also help to develop an innovation culture that permeates all areas, departments, and teams. Training also helps in spreading the AI knowledge across teams and enabling the organization to stay ahead of the market trends and experiment with the latest tech advancements. Through such literacy programs, the company can redefine itself as a more attractive organization.
A successful AI-powered organization also needs the right structure. Shaping the right organization is one of the most challenging problems a company can face. At NTT DATA we follow a hybrid model that combines the advantages of a Center of Excellence (CoE) and a Hub & Spoke approach. The Center of Excellence brings together all data and AI developments, defines the AI and Data Governance practices, and connects teams (e.g. Data Scientists, Data Engineers, and Data Analysts). The Hub & Spoke (H&S) consists of putting resources and technological know-how close to the business areas (Spokes) while the Hub enables the democratization of analytical capabilities.
Another important topic in the AI transformation journey is the opportunity for the organization to differentiate. But how to achieve differentiation from the competition when any company out there can access productized AI e.g. pre-trained industry-specific services such as Natural Language Processing or Computer Vision, etc.? Part of the strategy could be to not only pursue partnerships with leading big tech companies but also to adopt open-source AI frameworks that offer ready-to-use technologies and architectures. This would allow the company to harness these frameworks, use cloud AI services and open-source systems, and accelerate its AI developments. It could also introduce a new working model across the AI lifecycle: data scientists reuse standardized components and refocus from dull, repetitive tasks, to creating real business value through AI.
Once a company has adopted the culture and the technology, it is then necessary to think how they can use that combination to stand out from their competition. Companies should be looking not only to benefit from Cloud, AI, and open-source capabilities to improve efficiency, accuracy, and foster cost savings, but also to stand out and visibly differentiate their products from the competition through Data and AI-specific service and product design methodologies.
David Pereira is a telecommunications engineer from the University of Vigo and PDD from IESE Business School. Partner of the NTT DATA Technology Area, David leads NTT DATA’s Data and Intelligence in EMEAL as well as the global Artificial Intelligence Center of Excellence for the group.
[1] D. Simchi-Levi, J. Gadewadikar, B. McCarthy and L. LaFiandra, “Winning with analytics,” Accenture, 2015
[2] The ins and outs of becoming a data-driven organization — Telefónica (telefonica.com)
[3] The AI first Company, Ash Fontana, ISBN : 0593423089
[4] A Data Driven Company, Richard Benjamins, ISBN: 1912555883
Excerpt from 60 Leaders on AI (2022) — the book that brings together unique insights on the topic of Artificial Intelligence — 230 pages presenting the latest technological advances along with business, societal, and ethical aspects of AI. Created and distributed on principles of open collaboration and knowledge sharing: Created by many, offered to all; at no cost.
