Episode 1: Unleash Your Inner AI Hero

Foreword

Written by Ronald van Loon

CEO & Principal, Intelligent World

The GenAI Effect: Disruption, Opportunity and the Future of Business

We are witnessing a transformative era of advances in AI. This has been most notably marked by its mass consumer adoption enabled by the rise of accessible innovations in General Artificial Intelligence (GenAI). 

In its relatively short time in the spotlight, the GenAI landscape has evolved rapidly. Now, the focus has turned to how businesses can take advantage of its seemingly infinite potential.  

While current sentiment towards GenAI feels like cautious optimism, strategic adaptation and a collective effort towards responsible stewardship, the path to creating commercial value from this powerful technology is still being trodden. Despite this, adoption rates are nothing short of remarkable. According to Qlik’s Generative AI Benchmark Report, 79% of business leaders have already invested in a GenAI tool or project.

However, with great power comes great responsibility. The level of disruption caused by AI and GenAI will be significant across many industries, requiring an urgent and proactive approach to governance and risk management. At this point, most organizations don’t have robust policies in place on how they, and employees who may already be using it in their personal lives, interact with the technology.

The ethical and security concerns associated with AI and GenAI are abundant, demonstrated by the focus of President Biden’s AI Executive Order establishing new standards for AI safety, and the recent AI Global Safety Summit in London, which culminated in the Bletchley Declaration on AI Safety.

These developments underscore a global commitment to ensuring the safe and responsible development of AI. The emphasis on international collaboration, consumer protection and promotion of innovation reflects a balanced approach, aiming to harness the benefits of GenAI while mitigating its risks.

But even as governments consolidate their responses to innovation, GenAI is already having a tangible impact on how businesses operate. It's re-shaping how they view hiring and workforce skills, given its increasing use in functions like marketing, sales and customer care. At the same time, demand for data engineers, machine learning engineers and AI data scientists is rocketing, with emerging roles like prompt engineer also gaining prominence. It’s these workers who will shape the commercial response to AI innovation.

These changes appear to have been met with optimism and a strategic approach to investment, rather than inertia – 45% of executives indicate that recent advances in AI are driving an increase in AI investment, while 36% are investing in a formalized AI strategy.

That’s a good start, but action does not always deliver positive impact. This guide features perspectives from data and analytics leaders across multiple industries who are on their own AI journeys, providing tangible tips from those responsible for optimizing data processes with AI.

Introduction

Real AI experiences from the Data Front Line

The buzz around AI and its potential business value continues to build. However, amongst all the noise there is a definite feeling of information overload. So, it’s understandable that many data and analytics leaders remain uncertain of the technology’s real benefits and their road to adoption. They know that their data teams are fundamental to the success of AI initiatives, yet the path is less trodden and so yields fewer examples to follow.

Visionary Voices in AI brings together real experiences and tangible advice from data and analytics professionals that are on their own AI implementation journeys.

Read on to hear perspectives on maximizing the impact of AI and enhancing the role of the human. In addition to this, experts also weigh-in on the public versus privately trained models debate, discuss balancing innovation and risk with governance and ethics, and much more.

AI has been around for decades, but it’s experiencing a huge surge in popularity. This is down to many reasons including: advances in adjacent technologies, increased data availability creating improved algorithms, and a growing range of practical applications. While it is seen as a tool for competitive advantage, conversation and consideration as to where it can provide the best value continues. However, we are already starting to see how data and analytics teams are using – or plan to use – AI in different industries and for different purposes.

Act 1

AI’s positive potential

Transforming information to create valuable insights

It’s clear that both the private and public sector see AI as an opportunity to enhance the value of their products and services. And this, in turn, is helping them meet expectations, drive efficiencies and in some cases power ahead of the competition. 

“AI has the potential to revolutionize healthcare by improving diagnostic accuracy, enhancing patient outcomes, improving post-surgery care and reducing healthcare costs. It is already being used in image diagnostics for early detection of cancer. Robot assisted surgeries are already being performed, which can help in performing key-hole surgery and hence reducing the recovery time for the patients. In addition to this, AI can increase productivity and the efficiency of care delivery and allow healthcare systems to provide better care to more people.”
Deepa Tambe, Head of Reporting Technology, Barts Health NHS Trust

Mitul Vadgama, Senior Data and Analytics Strategy Manager at Lloyds Banking Group explains how the technology can bring value to the banking sector: “AI has the potential to transform financial services, such as banks, by providing personalized services, enhancing security, optimizing operations and enabling data-driven decisions. Adopting AI technologies can help banks stay competitive in a rapidly evolving industry, while delivering better services to their customers.”

Dave Elliott, Global Data and Analytics Manager at Mayborn Group, which owns global baby brand Tommee Tippee, agrees with the opportunities for improving services and meeting customer demand, but also for upgrading the back-end operations that enable this. He explained: “AI provides a vast array of opportunities… be that improving and optimizing the supply chain, enhancing the consumer experience through natural language processing or generative AI, or through core AI functions such as sales or demand forecasting.”

But this isn’t the only advantage that Mayborn Group sees from continued use of AI. Mark Little, Dave’s colleague and Principle Business Intelligence Specialist, explained: “We’ve used AI for a number of things. The one we have seen the biggest return from is sentiment analysis. Using the technology, we can look at all our reviews from across a number of platforms and get a general idea of how we are performing. Without it we’d have to spend hours collating and reading all the text.”

It's not just private organizations that see the potential for enhanced decision-making. Calum MacIver, Corporate Information Manager at The Health Informatics Service – hosted by the Calderdale and Huddersfield NHS Foundation Trust – highlighted how the NHS is benefitting from the predictive capabilities enabled by AI.

“Within the NHS, AI is becoming more and more visible in its use, from assisting diagnostics to predicting admission and attendances. The analysis we are seeing from the technology is opening up new avenues for research from all areas of the health service.”

Gaining an edge with speed

The theme of speed is very apparent across experts’ experiences with AI and when it comes to citing its value to businesses. Data and analytics teams play a fundamental role in turning disparate information into tangible business insights. With so much data available and the need for real-time insights, manual analysis can often be slow and insights are out of date by time of completion. It’s clear that AI can bring huge opportunities in this area.

As Priscila Papazissis, Data Product Manager at Localiza & Co., a car rental company headquartered in Brazil, notes: “AI is transforming access to data and analytics and promoting data driven decision-making in my organization. It brings about the opportunity to capture, treat and show huge amounts of data in faster ways than ever before thanks to machine learning algorithms. Now the time between the business event and collecting the results has reduced so we are informed faster and can make decisions that make a difference.”

John Delligatti, Director of Digital Supply Chain Transformation at SDI adds: “It's one thing if AI saves me, personally, a bunch of time, but if I can save 10 or a hundred people time – it's got tremendous value. So, I would say begin investigating these tools right away. Encourage your team members to use them.”

Embracing the unknown

Businesses continue to grapple with uncertainty due to the macroeconomic context, technology advancements and evolving regulation. The good news is AI can help build predictive models that allow you to anticipate future trends and outcomes. This equips data and analytics teams with the knowledge they need to help the business move forward confidently. It also frees up your teams’ time, so they can think strategically using the insight shared by AI.

“We can transform the roles of people who use and make decisions with data, because we can let AI do the repetitive work to identify issues. For example, identifying a fraud that lived inside the data would be near-impossible to find with only a human eye.” ‍
Priscila Papazissis

On the opportunity AI brings for large, legacy organizations such as the NHS, Calum MacIver explains that: “In terms of accessing data, AI has yet to play a big role, as we’re still relying on our time-tested processes to get at the data required. It’s more about what we look to do with that data once we get it. Looking for patterns, predictions and anomalies within the data are three areas we expect to make the greatest use of AI, more specifically Machine Learning.”

Deepa Tambe shares Calum’s view on the potentials of AI to predict unknowns: “In healthcare we collect tons of data from every interaction with patients and machines. What we need is the use of data for building insights. Predicting the pressure points, predicting the flow of patients in the emergency department and the resources needed to cater to the high demand of the hospitals are areas I believe it could really help. We are already using predictive analytics and alerting mechanism to notify operational teams for the correct bed type requirements. This was developed during the peak of the pandemic which hospitals found very useful.”

Define your needs before adopting 

With all hype around AI, it’s understandable that many might look to integrate it as quickly as possible. But doing so before assessing the challenges you want it to solve is risky. Guidance from our Visionary Voices was unanimous – adoption of AI for the sake of it is not impactful. Instead, the AI journey must start with a problem or challenge. Only then will you see the true value of the time and investment spent, and be able to take full advantage of its capabilities.

Mark Little notes: “AI will have a huge impact, but you need to make sure that you have a proper use case so that it comes back with a return. If you go in with no idea and just play around with the technology, you will probably conclude that it’s just a ‘trend’. But if you have a defined use case where you can measure the outcome, your experience of using AI will be much more positive.”

Martin Sahlin, Founder and CEO of Stretch Qonnect, which helps sports clubs with detailed, simplified and easy-to-understand analytics, adds: “Everyone is talking about AI, but many don’t know why or what to do with it. In order to get the right outcome and not to go down the wrong path and believe in something wrong, initiatives and investments in AI need to have a very clear and particular challenge in mind. What’s more, there needs to be a tangible measuring of the outcome.”

Organizations such as HCL Technologies identified a key challenge and have created solutions to proactively help the business.

“We have a data scientist team which works in Python and has created many ML models for business uses. One example is the Attrition Prediction Analysis – which helps us predict which employees are at risk of leaving the organization and allows us to support them accordingly.”
Rahul Gupta, Associate General Manager

Act 2

Re-defining the role of the human

New roles are forming, but not for all (yet)

A clear benefit of AI that data experts have already started to witness is the opportunity for it to uncover answers to questions they didn’t even know they had. This is thanks to its ability to analyze and process large volumes of data, demonstrate the relationship between data points and identify patterns and make predictions. This means business users and analysts don’t need data scientist-level skills to benefit from automated tools.

“AI enables me to better understand the relationship and context of data that I might not have noticed at first. It has also led me to question certain things that I wouldn’t have questioned before when analysing manually” - Michal Lecian, Business Intelligence Analyst at Dolphin Consulting.

Another key benefit of AI is the opportunity to create new, strategic roles and responsibilities. In doing so, data teams can step back from manual data wrangling and focus on analysis and insights gathering.

Filippo Orlando, Head of Advanced Analytics at Unieuro S.p.A, the Italian consumer electronics distribution leader said: ‘’AI enhances our team's capabilities by automating tasks, providing insights and improving efficiency. It's reshaping roles to focus on higher-level strategic tasks. In some cases, new AI-focused roles have emerged to manage and optimize AI systems.’’

Martin Sahlin agrees about looking at AI not as a replacement of jobs, but rather an accelerator of work: “You need to challenge AI. It’s important to be critical when looking at the results and monitor outcomes of AI very closely. This will be a bigger and bigger part of data roles going forward. Repetitive tasks can easily be outsourced to AI, leaving the more innovative work to people.”

But not all companies are ready for this shift or the addition of new expertise. Our Generative AI Benchmark Report highlighted that just over a third of organizations are planning to train data models fully in-house, while 60% are considering partially using third-party resources to do so, and only 4% are doing this fully in house.

Mario De Felipe Pérez, Chief Transformation Officer at Grupo ASV continues: “We are relying on specialized partners to deploy the technology. If it reaches sufficient maturity and relevance within the company, we will consider incorporating specialized personnel. What we are doing is training many people within the company in Artificial Intelligence. Mainly, we are working on how to use Generative AI to increase productivity in our call centres and in departments such as legal, financial or marketing.”

Act 3

Unleashing AI innovation safely

Achieving tangible AI benefits isn’t as easy as ‘just adopting.’ It also means balancing this innovation with the associated risks. This involves careful testing, regulatory compliance and closely monitoring AI-driven initiatives. It’s crucial to evaluate the potential benefits against possible drawbacks to make informed decisions about investment.

“The rise of Generative AI, such as ChatGPT, opens up opportunities and challenges across numerous domains. It encourages humans to think creatively about how Artificial Intelligence can enhance human endeavours, but it also underscores the importance of ethical use, responsible AI practices, and ongoing research to address the evolving landscape of AI applications. Managing innovation with ethical considerations will be important as we continue to leverage Generative AI and other advanced AI technologies moving forward.”
Mitul Vadgama

Priscila Papazissis added: “We have to start thinking and studying how to use this technology inside the organization in a secure way, with all the risks mapped, because in my opinion, there’s a lot of possible risks involved with AI.”

Approach publicly trained AI with caution

As an example of these considerations in action, many businesses have turned to publicly trained models in the race to adopt AI. This is mostly due to the time and expertise required to build bespoke, private versions. However, what is clear from the experts is that precautions must be taken before heading down this path.

Rahul Gupta, Associate General Manager at HCL Technologies explained: “The company has banned employees to use company proprietary information in ChatGPT and other similar models as there is a data security risk. This is because any data shared could be used to either train the model publicly.”

In the context of Generative AI, Dave Elliott focuses on the importance of challenging the data: “The rise of Generative AI comes with both excitement and caution. However, it does emphasize the need for organizations and users alike to ensure that foundational data knowledge and fluency is in place. Without this, we face falling into a scenario in which we follow the results without fully understanding the implications or being in a position to question. After all, to really leverage data for better decision making, no matter the source, we must be able to read, work with and challenge the data.”

‍The consensus, it seems, is for a healthy scepticism towards public AI models. Consistent concerns surround the ability to maintain data quality and understand potential biases. In addition to this managing noisy or incomplete data, handling ambiguity and adapting to evolving language were all noted as things to consider.

John Delligatti understands the value of creating and using your own AI: “When things are developed internally, or with the expertise of subject matter experts at your company, they’ll take longer but hopefully provide better results. Real concerns about the accuracy of AI models, predictions and generated answers comes from models that are publicly trained, not trained on your data, and frankly, where you can’t necessarily identify sources of statistics properly. We’ve all heard of examples where ChatGPT can pass a medical exam, but we’ve also heard of examples where ChatGPT attempts to draft a legal document and cites cases that don’t exist. So, you need to take everything that comes from these models with a grain of salt and fact check yourself.’’

Deepa Tambe is also concerned on the matter of using accurate data, stressing the full team using need to be educated on its importance: “As a basic fact the data which goes in is the data we get out. So, we need to focus on the accurate data being captured to give the right outcome. We still need to focus on basic principle of data accuracy and data literacy and educate people (of non-data & analytics background), the importance of it. We can only progress if we can bring everyone on the exciting journey of AI.”

Henri Rufin, Head of Data & Analytics at Radiall, shares his thoughts on the dangers of generative AI: “Using generative AI without any prior knowledge might be dangerous, as these technologies feed on the data you provide them which could easily become a security breach. Working with generative AI requires a deep understanding of the underlying technologies, along with a plan to avoid misappropriation or any ethical conundrum that may arise in the future.”

Reducing risk with data integration 

Data integration and management takes on new importance in the age of AI, alongside robust governance processes. It also requires a more flexible approach to data management, given how quickly an influx of new data inputs can alter an AI model.

“Organizations should aim to enhance data integration processes to support AI initiatives. They can do this by: implementing robust data governance, employing advanced integration tools, automating ETL pipelines, utilizing data lakes, ensuring real-time data streaming, maintaining data security and compliance, monitoring performance and fostering a culture of continuous improvement.”
Filippo Orlando

Dave Elliott explains how Mayborn Group is approaching data integration for AI, but also other technologies: “[Data integration] is a key foundation to all aspects of the organization’s data journey. As such, there is a core program to support the automation of data acquisition and integration underway. The focus of this program is to build a solid, governed and trusted source of data across all aspects of the organization to support both current data and analytics offerings, but also as a foundation to build upon new and emerging technologies.”

Prioritize governance and security

That’s not to say data governance is easy, given the volume of data being processed by businesses every day:‍

‘’We are very accustomed to supporting a structured data governance policy and ensuring its quality, but when we talk about unstructured data such as audio, video or text, the problem is much more complicated. We have little experience, and we are facing the problem of properly labelling unstructured data so that it can be processed by AI models.’’
Mario De Felipe Pérez

As part of data governance, data teams also face security challenges. Filippo Orlando shares what his team at Unieuro S.p.A. are focused on: ‘’I prioritize AI data privacy and security through encryption, access controls, and strict compliance with privacy regulations. This safeguards sensitive information, upholds ethical standards and builds trust with stakeholders.’’

Go forth and experiment 

However, this is where the balance between innovation and risk has to be struck. Upholding robust data governance and security shouldn’t put you off experimenting with AI – at the very least to understand where the most appropriate use cases are. If you start with setting clear objectives and ensuring data is properly anonymized and protected, that should make a good foundation for any experimentation.

In practice, this means cross-departmental collaboration and starting with a small, trusted group, according to Henri Rufin: “At Radiall, we learn by doing and we believe in proof of concept. AI is definitely something we want to investigate. While we must deal with extra care with generative AI, we can not stay idle and be scared by the danger these technologies hold. We have started an AI initiative which, I hope, will lead us someday to deliver new services to support data literacy and data governance across the company. We work closely with IT and security departments to minimize risk, but also by establishing a trusted group of people to experiment alongside you, before rolling something out to the public.”

Conclusion

Turning AI from hype to reality

It’s clear that AI presents an exciting opportunity for businesses. With the help of the technology they can make sense of data better and faster than ever before – empowering employees and enabling better outcomes for customers. Beyond this, it has the ability to change the way data teams work and enhance the value they bring to their organization.

As with all new technologies, something as game-changing as AI doesn’t come without doubt and speculation over its impact if used incorrectly, or used without being underpinned by data governance and ethics.

So, what learnings would our Visionary Voices share with their peers on how they can start taking advantage of AI now?:

Start small, test and learn:

“Ensure that data fluency within the organization supports understanding both the use of and the outputs of any AI driven processes. Start small, test and learn, don't be afraid to fail and ensure that the business understands that it is a learning journey.” - Dave Elliott

Stay committed to ethical and responsible practices

“AI is a journey, and the key is to start taking small steps toward harnessing its potential while continuously learning and adapting. By approaching AI strategically and with a commitment to ethical and responsible practices, you can unlock new opportunities and stay competitive in a rapidly evolving landscape.” - Mitul Vadgama

Never forget the importance of the ‘human touch’

“There are a lot of ideas out there from the AI experts which can be applied to innovate services for the care of patients. However I think the personal human touch of care should not be lost while running to improve the unknowns.” - Deepa Tambe

Start experimenting now

“My advice is that if you haven’t already, you have to get started. If you don’t plan to, you will be left in the dust by your competitors. AI isn’t the be all and end all but it’s a great tool and a great place to start experimenting. Start asking it questions specific to your business and see what it knows, start feeding it coding codes when you get an error and see what it says. These are the types of ways you can familiarize yourself – and encourage your teams to as well.” - John Delligatti

Work with expert partners

“First of all, you have to experiment with the technology to really understand the state it is in. Secondly, talk with specialized partners who can provide knowledge, experience and possible business cases. Thirdly, look for use cases through workshops within the organization.” - Mario De Felipe Pérez

Always keep humans in the loop

“AI can really help with out-of-the-box thinking. However, you have to be very careful about the results it shares – it isn’t always right. That’s why you should see it as an enhancement to the work being done by humans, rather than a replacement.” - Michal Lecian

Good, clean and governed data is crucial

“When it comes to AI technology, we have to remember that, no matter what, it's data based. So, data quality is fundamental. Nothing will work properly if your data is not properly governed. That means you have to focus on data literacy. At Radiall, we have been focusing on data integration, dashboard automation and data quality processes before even thinking about taking advantage of AI. That being said, I strongly believe AI, and Generative AI in particular, is going to change the way users interact with their data for the better, thus supporting and improving data literacy in the long run.” - Henri Rufin