The Journey to AI

Part 1:

Scaling Data to Seize the Possibilities of AI

Artificial intelligence has been called the new electricity, and from voice assistants to recommendations on streaming sites, AI is already powering many day-to-day interactions.

It’s making waves across the business world, too. Companies are under pressure to transform how they operate, interact with partners, and engage with customers. As such, AI – with its deep learning and data analysis capabilities – is becoming a necessity. And Covid-19 has shown it is one of the most important innovations organizations can employ to sustain operations experiencing unprecedented pressure.

Many companies are already switched on to this –

And the pandemic has put many organizations years ahead on their digital transformation roadmaps.

But deploying AI at scale continues to present challenges, such as insufficient data, data complexity, a lack of talent , and distrust of AI systems.

As a result,

8 out of 10 AI projects fail or don’t get off the ground.

However, the incredible business potential of AI solutions in predicting future outcomes and automating decisions and processes cannot be ignored.

“The insights provided by Big Data and AI will fuel the next wave of innovation. This will allow leading-edge companies to remain competitive and stay ahead of the game.”

Richard Wilkins, Distinguished Engineer, IBM Cloud and Cognitive Software

So what do companies need to do to build the foundations of a successful AI-fueled future?

The answer lies in getting data in order and combining the intelligence of AI and the agility of the hybrid cloud to scale its impact.

Accelerated Roadmaps

The impacts of Covid-19 – such as spikes in customer service volume or broken or uncertain supply chains – have helped many organizations to see the true potential of data and AI. Almost overnight, the pandemic pushed them into a digital-first strategy, and 59% of enterprises surveyed by IBM said that the crisis had accelerated their digital transformations.

And while AI adoption has remained nearly flat over the past year, the momentum is shifting and companies are planning significant investments in this area:

The top areas of AI companies are planning to invest in are data privacy, automation of processes, customer care, virtual assistants, and business process automation, according to IBM’s Global AI Adoption Index 2021, which surveyed more than 5,000 global businesses. And according to 90% of the IT professionals polled, the ability to access data from anywhere is key for increasing AI adoption.

Across industries, AI and cloud have risen into the top three technologies ranked according to revenue impact during the pandemic.

“During the crisis, data and AI helped businesses auto-pilot operational decision-making. They helped enterprises to operate efficiently and supported hyper-scale growth.”

Anup Kumar, Chief Solutions Architect, Data & AI, IBM Asia Pacific

Companies that have this digital ecosystem in place, with the ability to connect any data and any cloud, are primed to take advantage of the new levels of effectiveness and growth opportunities offered by predictive and AI technology. It will also help them reduce regulatory and reputational risk.

AI has become an urgent strategic imperative that will distinguish a successful business from its competitors.

Barriers to Adoption

For any company embarking on its AI journey, there are challenges that must be addressed. More than a third of IT professionals surveyed for the Global AI Adoption Index reported their business has not deployed any AI projects.

Top 3 barriers

Limited AI expertise or knowledge


Increased data complexity and data silos


Lack of tools or platforms for developing AI models


Compounding these issues, 81% of business leaders do not understand the data and infrastructure required for AI.

For Wilkins, it’s access to quality data that is both at the heart of this issue and the key to solving it. “There are many aspects to operationalizing AI and having too much of the wrong data is just as bad as not having enough data,” he says. “The trick is ensuring you have enough data in the context of the problem that you are trying to solve.”

Transformative Impacts

As businesses look to the post-pandemic future, it will be crucial to give data scientists and developers the capabilities they need to overcome these issues and scale AI. With a majority of executives telling another IBM survey, on the business value of AI, that they anticipate ongoing business turmoil, this will be vital.

Smarter organizations, fueled by data, are using AI and the agility of hybrid cloud to transform business models and create innovative, secure, and personalized digital experiences.

During the pandemic, organizations that had already meaningfully built technology into their operations consistently outperformed their peers in revenue growth by six percentage points on average.

As businesses become more familiar with the potential of AI, automation, for example, is one technology that is becoming more deeply embedded in day-to-day operations. In fact, 80% of companies are already using automation tools or plan to in the next 12 months. And over the past year, many organizations of all sizes either adopted natural language processing or recognized its value.

Enabling Insights

Those in the know understood this potential well before the current crisis. As Kitman Cheung, Technical Sales Leader – APAC Data, AI and Automation – IBM Technology, says: “Researchers, technologists and scientists have long recognized the potential of these technologies to drive profound changes in our lives.”

And in practice, these tools can have a powerful impact, helping businesses with everything from streamlining processes to gaining important insights into their business and improving customer satisfaction.

Data lake systems, for example, let companies store vast amounts of raw data from a variety of sources until it is needed. Allowing organizations to glean insights from diverse data sets, they provide a platform for real-time analytics and more agile data-driven decisions – enabling competitive advantages including a deeper understanding of customers and better responsiveness to trends.

“The ability for AI to discover additional insights in existing data sets will be key for companies to innovate and provide new services that both retain and expand their customer base.”

Richard Wilkins, Distinguished Engineer, IBM Cloud and Cognitive Software

AI-powered software, such as IBM Cloud Paks, can help enterprises build, modernize, and manage applications securely across any cloud. And then there is machine learning. IBM Watson – an open, multi-cloud AI platform – for example, is helping organizations drive smart reinvention of workflows and technology to transform into cognitive enterprises. They can tap into unstructured data, improve decisions based on real-time trends, and quickly and securely build cognitive applications.

With this technology, organizations can integrate AI into their existing workflows to predict and shape future outcomes, automate decisions, experiences, and processes, and optimize employees’ time to focus on higher-value tasks. It is already at work across industries, doing jobs such as keeping millions of elevators moving, predicting when they will break down, and proactively fixing them. By powering virtual agents trained on customer enquiries, it is also enabling businesses to help customers faster.

Superior Outcomes

Such advances can be transformative for businesses, driving greater efficiencies, saving costs, and empowering employees with AI-driven insights. During the pandemic, AI adoption has become positively correlated with superior business outcomes in revenue, cost, and profitability. This is true across industries and regions.

The potential rewards are only going to grow. It’s predicted AI could add almost $16 trillion to the global economy by 2030. And with the onset of 5G – which will allow data to travel and reach devices nearly instantly – data processing potential will increase. This will further boost the opportunity to harness the capabilities of AI at scale.

Part 2:

Unlocking the Power of AI

Of course, AI relies on data – and making sense of it is vital if companies are to reap the plentiful rewards AI and predictive technologies offer. It can be a complex task – especially if that data is 30 years of engineering and drilling knowledge buried in unstructured documentation.

Efficient use of engineering records brought cost savings of AUD 10 million.

That was the challenge facing Australian oil and gas company Woodside Energy. Unable to effectively extract the meaningful insights it needed to make fact-driven decisions on complex projects, the organization brought its problem to IBM.

With IBM Watson and cognitive computing, Woodside was able to draw on wisdom accumulated by thousands of engineers over decades. About 33,000 documents including testing results and reports were uploaded, and these were scanned by text analysis and machine learning algorithms for correlation. The AI was then trained to think like a Woodside engineer

This put the data at the fingertips of employees across the organization – to great effect. Efficient use of engineering records brought cost savings of AUD 10 million. And there was a 75% reduction in the time teams needed to search through and read data sources.

Collecting and organizing data like this to inform real-time insights and decision-making will be fundamental to companies taking advantage of the possibilities of new technologies.

“Those that leverage AI and predictive technology to augment decision-making will outperform competitors that fail to capitalize on AI.”

Kitman Cheung, Technical Sales Leader, APAC Data, AI and Automation, IBM Technology

Establishing a foundation of business-ready analytics in this way will be essential for future growth.

A Foundation for Success

The uncertainty caused by the pandemic has made the levels of resiliency and efficiency realized by Woodside Energy even more crucial.

AI is a multifaceted technological innovation with layers of interconnected and moving parts, however, and embracing it during a period of flux such as the current crisis can be complex. It has been described as being like “changing out a jet engine while the plane is flying through turbulence at 35,000 feet”. But it doesn’t have to be such a bumpy ride.

A crucial element of AI success is building an efficient, agile data architecture through a hybrid multi-cloud strategy that ensures data can be connected to any cloud, anywhere.

“Such a solid analytics foundation is like a vaccine against various uncertainties in business. It provides immunity during unfavorable situations and boosts readiness to respond quickly to new opportunities for growth.”

Anup Kumar, Chief Solutions Architect, Data & AI, IBM Asia Pacific

Building a Strategy

Regardless of where a business is on its journey to AI, breaking an AI strategy down into manageable pieces can serve as a guiding principle. One example of such an approach is a framework developed by IBM called the AI Ladder.

This advises that organizations focus their transformation on four main areas: how they collect, organize, analyze and ultimately infuse data, machine learning and AI throughout their business.

1. Collect

AI is only as good as the data that feeds it, so once an organization has modernized its architecture, it’s important to make its data simple and accessible. Hybrid data management products can help manage structured and unstructured data on premises and in the cloud.

2. Organize

Confidence in AI relies on data that is trustworthy, complete, and consistent. So it must be cleansed, organized, catalogued, and governed to ensure only authorized individuals can access it. IBM Watson Knowledge Catalog enables this, letting users quickly find, curate, categorize, and share data.

3. Analyze

Once data is organized in a trusted, unified view, AI models can be built and scaled across a business. This allows companies to gain insights from all of their data, no matter where it resides, and engage with AI to transform their business.

4. Infuse

Then it is time for AI to be put to work in multiple departments and within various processes – from payroll, to customer care, to marketing – drawing on predictions, automation, and optimization.

By using the AI Ladder as a guiding framework, enterprises can lay the foundation for a governed, efficient, agile, and future-proof approach to AI. A robust data-processing foundation such as this will be pivotal for businesses to achieve AI capabilities in the years ahead.

When leveraging the cloud and an open-source foundation, a data fabric can deliver the scale and compute power needed for digital transformation.

But before this can happen, organizations must recognize that they need a systematic approach to ensure AI and machine learning can be trusted and operationalized. Aspects of trustworthy AI include fairness, robustness, privacy, explainability and transparency. Operationalizing trustworthy AI requires the bringing together of people (expertise), process (best practices), and platform (technology).

Companies with effective data strategies are also using data fabric to help make data visible and usable across the enterprise. A data fabric is an environment that includes an architecture and set of unified data services, which together support consistent data capabilities across an organization’s network – both on their premises and on multiple cloud environments. This helps businesses put their hands on the right data just in time, at the optimum cost, with end-to-end governance – regardless of where the data is stored.

The governance, security and regulatory compliance built into the fabric consistently across all data creates trusted outcomes for AI initiatives. Because a data fabric eliminates the need for independent tools that must be manually integrated, there are cost and operational efficiency benefits too.

“It is simply not possible to operationalize AI at any scale without a robust data strategy.”

Anup Kumar, Chief Solutions Architect, Data & AI, IBM Asia Pacific

Part 3:

An AI Ecosystem

Successful scaling of AI projects can be a tricky business.

Broadly, what’s needed here is a step change in the role of AI, enabled by a shift in mindset that views the technology not as an experiment but as a strategic capability to be embedded throughout the business.

This kind of thoughtful and holistic embracing of AI roots the technology in innovation and competitive differentiation. Then it deeply integrates it into operating models and workflows, organizational structures – and even in cultural values and ethics. The workforce, meanwhile, is empowered with AI-driven insights that helps it make real-time decisions.

“Cognitive enterprises are fueled by data and they create value when their knowledge workers can leverage data to deliver better business outcomes”

Kitman Cheung, Technical Sales Leader – APAC Data, AI and Automation - IBM Technology

Trust is important, too – 90% of companies using AI say their ability to explain how it arrived at a decision is critical. So it needs to be operated with robust ethical principles and rigorous operations and governance. This is where a data fabric that consistently delivers trusted, secured data can bring value.

In the years ahead, the speed, agility, and resilience AI brings will be essential for companies to maintain a competitive edge and protect themselves against periods of uncertainty or shocks. As a result of the pandemic, 84% of executives said they expect a steady or increased level of organizational focus on AI, according to the IBM Institute for Business Value. So it seems many companies are on the right track.

Meeting New Expectations

That increased level of focus will be vital.

Another effect of the pandemic companies must navigate is that it has changed how customers interact with businesses. It has raised their expectations, too. As such, another big theme in the coming years is going to be customer retention and loyalty.

Personalization, underpinned by AI, will be key to keeping customers happy and ensuring they come back. And once again, robust data foundations are crucial to the systems that will enable this.

For retailers with bricks-and-mortar operations, the sudden shift to e-commerce forced by lockdowns made an increased understanding of customers to provide enriched, personalized experiences vital. The Mall Group, a leading department store and shopping center developer in Thailand, already had ambitions to build new experiences and become an omnichannel retailer. The reduced sales volumes spurred by the pandemic made it essential to accelerate the process.

It is using IBM technology to do this by better understanding changing consumer behavior and trends and glean real-time insights from its shopping databases, which include millions of loyalty card users. For example, it can now draw on its historical data to offer a more personalized service, by offering, say, limited-edition trainers to a customer it can see is likely to purchase shoes from that brand.

The Mall Group’s adoption of advanced analytics and AI technology has improved the speed of its data analytics capability by 60 times, and it has given it the ability to work seamlessly across platforms. The company says this is a step toward providing a full-scale omnichannel shopping experience in the future.

IBM Watson works in the background in real time to determine what is being discussed

Intelligent Assistants

It’s not just retail seeing this kind of impact, though. AI technologies, including virtual agents, are transforming customer service across industries. Their ability to understand human sentiment and interact naturally with people can bring powerful results.

In New Zealand, an AI agent called Tala – powered by IBM Watson – is helping to break down consultation barriers to get important feedback from one of the country’s ethnic minority communities. Developed and designed by consultancy firm Beca, the natural language agent interacts in real time in Samoan or English, helping authorities interact with communities for whom English is a second language and that might not engage with more traditional forms of consultation, such as scripted surveys.

It gives people a chance to participate when they want, and in a place and language they are comfortable with. IBM Watson works in the background in real time to determine what is being discussed – for example local facilities – and ascertain sentiment. This information is then displayed on a visual dashboard to give decision-makers an overview of attitudes in the community.

“We use a range of IBM Watson services to collect, process and analyze information in real time, using AI to improve efficiency and remove the need for human intervention. This gives great scale and speed at great value.”

Matt Ensor, Business Director – Advisory, Beca

Ensor explains further: “For example, for one large organization, it took just a few days to obtain and digest detailed information on why some members supported an initiative and why others didn’t, and make this information available in real-time interactive visualisations. It would have taken weeks or months to do this using traditional methods.”

More than 500 people were involved in training Tala, which enabled it to learn how to have a natural conversation in multiple languages, on the phone and online.

Predictive Healthcare

Interacting with people in this way has become more important during the pandemic, which made in-person meetings impossible in many circumstances. Another industry that has been affected by this is healthcare.

But the strain on the sector has also propelled innovation through AI applications that provide greater ground cover for staff, introduce process efficiencies, and assess and even treat patients without straining resources or increasing operational costs.

iKure has helped millions using AI and IBM Cloud across 7 Indian states, 9 African countries and Vietnam.

In India, where the majority of people live in rural areas without immediate access to basic healthcare, social enterprise iKure is using AI in a predictive model for cardiac care. The system uses machine learning to rank diagnostic test data collected by health workers from its clinics, which is then transmitted to iKure’s platform, created on IBM Cloud, to determine the most at-risk cases for analysis.

So far, iKure has helped millions of people across seven Indian states, and its platform has also been implemented in nine African countries, as well as in Vietnam. Its next goal is to scale its operation to 200 hubs that will reach 25-30 million people in India.

Remarkable Potential

During the pandemic, businesses have had to develop new strategies to continue to meet the needs of customers while cutting costs, becoming more responsive and making faster, more informed decisions. Companies that can overcome the barriers to adoption and deployment will be able to deliver value from AI in the years ahead.

Leading data, AI, and integration capabilities infused with a hybrid cloud strategy are key to companies truly taking advantage of the remarkable potential of this technology, accelerating the digital journey and providing exponential opportunities for growth.

Ultimately, the successful adoption of AI will augment human capabilities and performance, creating better outcomes for customers, employees, partners, and other stakeholders, too.

Scale AI that runs anywhere and integrates everywhere

IBM is the only technology partner with the AI Ladder. This framework simplifies and automates your AI journey, and brings together leading data, AI, and integration capabilities on our open, hybrid cloud platform.

Find out more
This page was paid for by IBM . The editorial staff of CNBC had no role in the creation of this page.