TagArtificial Intelligence

Analytics and AI: Humans and Machines are Good at Different Aspects of Prediction

Mostly driven by growth in the IoT, and the widespread use of internet, social media and mobile devices to perform search, send text, email and capture images and videos, the amount of data that we are producing on a daily basis is startling.

Consequently, companies are turning to data analytics and AI technologies to help them make sense of all the data at their disposal, predict the future, and make informed decisions that drive enterprise performance.

Although adoption of analytics and AI systems is increasingly extending in more mission-critical business processes, the implications of these emerging technologies on busines strategy, management, talent and decisions is still poorly understood.

For example, the single most common question in the AI debate is: “Will adoption of AI by businesses lead to massive human job cuts?”

Borrowing lessons from historical technological advances, yes, certain jobs will be lost, and new ones also created. However, machines are not taking over the world, nor are they eliminating the need for humans in the workplace.

Jobs will still be there albeit different from the traditional roles many are accustomed to. The majority of these new roles will require a new range of education, training, and experience.

For instance, nonroutine cognitive tasks demanding high levels of flexibility, creativity, critical thinking, problem-solving, leadership, and emotional intelligence do not yet lend themselves to wholesale automation.

Analytics and AI rely on data to make predictions

As more and better data is continually fed to the machine learning algorithms, the more they learn, and improve at making predictions.

Given these applications search for patterns in data, any inaccuracies or biases in the training data will be reflected in subsequent analyses.

But how much data do you need? The variety, quality and quantity of input, training and feedback data required depends on how accurate the prediction or business outcome must be to be useful.

Training data is used to train the predictive algorithms to predict the target variable, while the feedback data is used to assess and improve the algorithm’s prediction performance.

Undoubtedly, advanced analytics and AI systems are only as good as the data they are trained on. The data used to train these learning algorithms must be free of any noise or hidden biases.

You therefore need to understand how predictive technologies learn from data to perform sophisticated tasks such as customer lifetime value modeling and profitability forecasting.

This helps guide important decisions around the scale, scope and frequency of data acquisition. It’s about striking a balance between the benefits of more data and the cost of acquiring it.

Humans and machines both have shortcomings

In the context of prediction, humans and machines both have recognizable strengths and weaknesses.

Unless we identify and differentiate which tasks humans and machines are best suited for, all analytics and AI investments will come to naught.

For instance, faced with complex information with intricate interactions between different indicators, humans perform worse than machines. Heuristics and biases often get in the way of making accurate predictions.

Instead of accounting for statistical properties and data-driven predictions, more emphasis is often placed on salient information unavailable to prediction systems.

And, most of the times, the information is deceiving, hence the poor performance.

Although machines are better than humans at analyzing huge data sets with complex interactions amidst disparate variables, it’s very crucial to be cognizant of situations where machines are substandard at predicting the future.

The key to unlocking valuable insights from predictive analytics investments involves first and foremost understanding the definite business question that the data needs to answer.

This dictates your analysis plan and the data collection approaches that you will choose. Get the business question wrong, conclusively expect the insights and recommendations from the analysis to also be wrong.

Recall, with plentiful data, machine predictions can work well.

But, in situations where there is limited data to inform future decision making, machine predictions are relatively poor.

To quote Donald Rumsfeld, former US Secretary of Defense:

There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don’t know. But there are also unknown unknowns. There are things we don’t know we don’t know.

Donald Rumsfeld, former US Secretary of Defense

Thus, for known knowns, abundant data is readily available. Accordingly, humans trust machines to do a better than them. Even so, the level of trust changes the moment we start talking about known unknowns and unknown unknowns.

With these situations, machine predictions are relatively poor because we do not have a lot of data to ingest into the prediction model.

Think of infrequent events (known unknowns) that occur once in a while, or something that has never happened before (unknown unknowns).

At least for infrequent events or happenings, humans are occasionally better at predicting with little data.

Generally so because we are good at comparison and applying prudent judgement, examining new situations and identifying other settings that are comparable to be useful in a new setting.

We are naturally wired to remember key pieces of information from the little data available or the limited associations we have had in the past.

Rather than be precise, our prediction comes with a confidence range highlighting its lack of accuracy.

Faced with unknown unknowns, both humans and machines are relatively bad at predicting their arrival.

The simple truth is that we cannot predict truly new events from past data. Look no further than the current Brexit conundrum.

Nobody precisely knew the unintended consequences of the UK leaving the EU. Leavers and Remainers both speculated as to what the benefits and disadvantages of leaving the EU maybe.

Of course, nobody knows what will happen in the future but that doesn’t mean we can’t be prepared, even for the unknown unknowns.

In their book Prediction Machines: The Simple Economics of Artificial Intelligence, Ajay Agrawal, Joshua Gans, and Avi Goldfarb present an additional category of scenarios under which machines also fail to predict precisely – Unknown Knowns.

Per the trio:

Unknown knowns is when an association that appears to be strong in the past is the result of some unknown or unobserved factor that changes over time and makes predictions we thought we could make unreliable.

PREDICTION MACHINES: THE SIMPLE ECONOMICS OF ARTIFICIAL INTELLIGENCE

With unknown knowns, predictive tools appear to provide a very accurate answer, but that answer can be very incorrect, especially if the algorithms have little grasp of the decision process that created the data.

To support their point of view, the authors make reference to pricing and revenue analysis in the hotel industry, although the same viewpoint is applicable elsewhere.

In many industries, higher prices are analogous to higher sales, and vice versa.

For example, in the airline industry, airfares are low outside the peak season, and high during peak seasons (summer and festive) when travel demand is highest.

Presented with this data, and without an understanding that price movements are often a function of demand and supply factors, a simple prediction model might advocate raising different route airfares to sell more empty seats and increase revenues. Evidence of causal inference problems.

But, a human being with a solid understanding of economics concepts will immediately call attention to the fact that increasing airfares is unlikely to increase flight ticket sales.

To the machine, this is an unknown known. But to a human with knowledge of pricing and profitability analysis, this is a known unknown or maybe even a known known provided the human is able to properly model the pricing decision.

Thus, to address such shortcomings, humans should work with machines to identify the right data and appropriate data analysis models that take into consideration seasonality and other demand and supply factors to better predict revenues at different prices.

As data analytics and AI systems become more advanced and spread across industries, and up and down the value chain, companies that will progress further are those that are continually thinking of creative ways for machines to integrate and amplify human capabilities.

In contrast, those companies that are using technology simply to cut costs and displace humans will eventually stop making progress, and cease to exist.

Talking About the Risks of AI and Cognitive Technologies

According to the recently published PwC’s 22nd Annual Global CEO Survey, 85% of the surveyed CEOs overwhelmingly agree artificial intelligence (AI) will have a significant impact on their business within the next five years. For this reason, they have plans to pursue AI investments.

This is despite the fact that the information gap between the data CEOs are requiring to make informed decisions and what they are getting from their teams has not closed. Lack of analytical talent, data siloing and poor data reliability are the primary reasons the data they receive is inadequate.

Nonetheless, the application of AI and the underlying cognitive technologies such as machine learning, computer vision, natural language processing (NLP), audio and signal processing, speech recognition, predictive systems and robotics are wide-ranging, with the potential to improve performance in nearly any activity that generates large amounts of data.

Highly-powered algorithms which are the basis of these computer systems are presented with large amounts of data and subjected to supervised, semi-supervised, unsupervised, reinforced and deep learning.

The goal is to train the algorithms to identify relationships or patterns between the inputs and the outputs and use those rules to predict future outcomes with input data alone.

For example, in healthcare, AI is being used to study patient clinical data and recommend diagnoses. In finance, machine learning algorithms are being used to analyze transactions and uncover fraud and money laundering.

In the retail industry, predictive algorithms are being trained to automatically group customers into various categories based on their needs or buying patterns. These insights are then used to prioritize sales efforts and tailor promotions.

In other cases, companies have piloted NLP technology to monitor social media sentiment. The technology automatically identifies conspicuous topics of consumer conversations and sentiment surrounding those topics.

The generated insights are being used to influence decisions on improving marketing and customer service.

It is no surprise then that CEOs are now exploring how to implement these new technologies in their business.

Hype-driven or well-informed investments?

As much as AI is a source of significant business opportunities, the same technology is also a source of significant threats that must be evaluated. This is essential for helping leaders make informed and intelligent investment and risk decisions.

It is foolhardy for leaders to jump on the AI bandwagon and expect to capture the promises of AI and cognitive technologies if they lack an understanding of whether, how, and where to invest in applying these technologies.

When almost everyone is talking about the opportunities of AI and cognitive systems, it’s easy to cave in to hype-driven or ill-informed investments and overlook the fact that AI and cognitive technologies are not the solution to every business problem or situation.

That is why it is critical to evaluate the business case for investing in these technologies and assess the potential impact on your company’s business model, culture, strategy and sector.

Take a holistic view of your business processes, products and markets to weigh where the use of AI maybe be practical, profitable and crucial.

Algorithms are only as good as the data they learn from

Given that AI capabilities are data-driven, closing the information and talent gaps is key to unlocking AI’s potential. AI-powered algorithms improve over time through their experience of using data.

They learn relationships between variables in historical data sets and their outcomes. The relationships are used to develop models, which in turn are used to used to predict future outcomes without needing to be explicitly programmed by a programmer.

The systems change and evolve depending on the data that is fed to the algorithms. This therefore requires the data that is fed to the systems to be accurate, complete, diverse, and free from errors and bias. If the data is incomplete, error-prone or contains innate bias, the algorithms are likely to display false patterns as well as magnify the bias leading to misleading outcomes that have far-reaching repercussions.

Since AI and cognitive technologies deliver outcomes based on historical or existing data presented to them, leaders need to acknowledge that these systems will not necessarily provide flawless outcomes.

That is why it is critical to have appropriate data governance structures and talent in place to monitor where and how these technologies are deployed across the organization.

Skilled personnel play the critical role of overseeing biases and risks emanating from algorithms. For example, these people help identify and mitigate risks associated with programming errors.

Understand the black box of AI

As business leaders lay the foundation to pursue AI investments and entrust key decision making processes to intelligent machines, it is worthy to demystify the ‘black box’ of AI.

This is the notion that we can understand the inputs and outputs of an AI-powered system, but don’t understand what happens inside.

Accountability is an important element of decision making, and in order to make AI systems accountable for their decisions, AI-based decisions need to be explainable in order to be trusted.

Rather than blindly entrust machines to make important decisions, leaders therefore need to develop an understanding of how the technology works and how it makes decisions.

Thus, business leaders must be able to identify and explain the layers of decision making which underpin the operation of the systems and influence the final outcomes.

For example, are you able to identify and explain which connections have predictive value in the multilayered deep neural networks? Although it’s impossible to analyze all the connections in a deep neaural network, it’s important to prioritize what you need to know, what you want to understand, and why.

Over time, through testing and measuring, or trial and error, you will be able to understand the thought process behind algorithms, trust the decisions they make and ensure a robust governance structure is in place to monitor these technologies as they mature.

Surge in Cyber Attacks

Big data has been a boon to the development of AI and cognitive technologies. Thanks to advances in technology, our digital lives are producing staggering amounts of data each day.

As a result, interest in AI application is surging as decision makers try to make sense of all the data at their disposal.

Nevertheless, leaders need to be aware that the more data is generated the higher the probability of cyber criminals or hackers targeting the company’s AI systems to steal personal data or business confidential information.

A major data breach can have unintended consequences that can create legal, brand and public relations issues for the business.

Therefore, as leaders seek to capture the opportunities of AI and cognitive technologies, they mustn’t turn a blind eye to the limitations of these systems.

They must also consider the various ethical, moral, and legal issues associated with the AI systems that their organizations deploy.

A Practical Approach to Using Artificial Intelligence for CFOs – Part IV

Part IV Getting After It: Take the Next Step and Make Your Investment in AI

If you haven’t had a chance to read Part I – Leveraging AI in the CFO SuitePart II – The Benefits of AI and What You Will Need to Make It a Success and Part III Where to Invest in AI, How to Measure the Financial Impact and Select Projects yet, please do so before continuing on.

There are four major investments you’ll need to make to use AI successfully in your business.

1. Develop an AI Strategy: This investment is about learning how to apply AI to your activities and selecting your best course of action. Consider using outside experts to help augment your thinking in this area if you are just starting your AI journey.

  • The first step of strategy development includes learning about AI, determining how it will be applied to the CFO responsibility areas, assessing the value of AI application for those areas.
  • The second step is to gauge the data needs (availability, accuracy, volume) and the cost of “creating” data that can generate the output required. “Quality, effectiveness, efficiency and insight are the four key pillars that really make this valuable stuff…” according to Nick Frost, KPMG Audit Technical Lead Partner.¹ Watch for these characteristics in your data. If they aren’t present, be wary of how you use your final product.
  • Using the value noted in a. above and the cost determined in b. an AI Strategy targeting the areas where AI will have the most impact can be constructed.
  • Skill/System assessment and timeline. Determine where growth in skills and systems are needed. The scope of these needs will also help create the resources required and a timeline. From a risk perspective, consider starting small (high expected return, low initial investment) and allow for greater investment as success is realized.
  • Include a change management plan to assist employees and other stakeholders in understanding the strategy and the impact it will have on them.

2. AI Software Selection: The investment in software will include the cost of the software and the expenses of the internal and external team members working on the process.

  • Use your Strategic Plan to target AI vendors that serve the areas highest on your list.
  1. On premise or cloud solution
  2. Data storage costs
  3. Integration with current systems.
  • If AI is new to you stay small and focused on high return, bite-sized efforts you can learn from.
  • Use your network to validate claims made from vendors in terms of system results, implementation timeline and cost.
  • This investment will include the direct payments for the software and internal costs for the selection team to do their work.
  • Our “AI Capital Investment Analysis” tool will help you summarize and communicate your planned investment in AI.

3. Implementation to Operation: It is important to focus the cultural change required during this stage to create an environment that craves the new learning AI brings to the table. The combination of our team’s desire to use AI wisely and a sound AI system add up to success. If either is missing, there is a good chance your implementation will fail.

Here are the implementation steps:

  • Research and mitigate the risks related to the implementation and data management.
  • Train and hire the skills to manage the system and leverage the new capabilities created by the AI.
  • Identify and manage the risks that are likely to occur because of the implementation.
  • Procure and implement the technology that fits your strategy.
  • Monitor and adjust the AI inputs and outputs to create optimum value for the AI stakeholders.

4. Ongoing AI growth: Your AI strategy document is the road map that will be used to plan AI follow up. It is a living document that requires updating.

  • Manage the ongoing operating costs of the AI system
  • Implement AI applications per the Strategic Plan
  • Change the priorities in the Strategic Plan as necessary
  • Consider new applications (see 1 above)
  • Assess current operating AI systems for optimization annually.

Artificial Intelligence holds great promise for financial professionals. It’s a key ingredient to enhancing the business partnering momentum established in the new millennium. Creating our AI Strategy, securing the skills to choose, implementing and operating AI systems, and growing these capabilities are new challenges demanding the attention of the CFO. Developing more efficient and “smart” transaction systems while improving decision support activities are huge value drivers for businesses today. Our ability to harness the power of AI to these means will be a significant measure of our success.

We’d love to hear about your AI experience (email us at info@erpminsights.com)!

¹Eleanor O’Neill, “How is the accountancy and finance world using artificial intelligence?” CA Today, July 31, 2016

A Practical Approach to Using Artificial Intelligence for CFOs – Part III

Part III Where to Invest in AI, How to Measure the Financial Impact and Select Projects

If you haven’t had a chance to read Part I – Leveraging AI in the CFO Suite and Part II – The Benefits of AI and What You Will Need to Make It a Success yet, please do so before continuing on.

Where the CFO can invest in AI to create a positive impact.

Now that we know what AI is and its benefits for finance, how can a CFO develop a plan around how to apply it in their business? To borrow a phrase from Stephen Covey, Begin with the end in mind. Visualize where you want to be and work backwards, considering what is preventing you from realizing your future today. This step will help prevent you from building AI around current systems and processes that are encumbering your digital transformation.

The next step in identifying where to invest in AI is to summarize the outputs your team creates for the company’s stakeholders. Define output as anything your team delivers to a stakeholder that they use. Examples of outputs include; invoices to customers, financial reports to management, pay checks/stubs to employees, borrowing base to the bank, work papers to the auditor, KPIs to the Board of Directors, credit information requests from vendors, accounts receivable aging report to the credit department, new project investment analysis for the CEO, productivity reports for the COO, etc.

​To be highly effective the implementation of AI is a multi-discipline exercise that will require resources from many parts of the business. A good example of this can be illustrated when using AI to assist in auto invoicing and payment applications. The sales department, manufacturing and shipping departments will provide data that allows these two functions to operate autonomously. The data from these departments will be incorporated into algorithms that function to determine how much, when and to whom to send an invoice; and, how to apply payments when the bank reports them as received.

​ Below are some important criteria to think about when selecting where to apply AI:

​ 1. Stakeholder focused; Serve your most important constituents first – Customers, Vendors, Employees (including management) and Directors

​ 2. Determine where AI has the largest potential impact

  • ​ Where improvements speed, accuracy and/or volume have significant impact
  • ​ Revenue generation
  • ​ Cost savings

 3. Understand the complexity of AI application.

  • ​ Data requirements
  • ​ System requirements
  • ​ Process requirements

Measuring the (financial) benefits of an investment in AI for a business

​Just like any other business case development, it is important to measure the benefits of investing in AI technology. These benefits are either tangible or intangible. Tangible benefits are those that can easily be quantified, you can put a value against. On the other hand, intangible benefits are difficult to quantify, but expected to occur as a result of the investment.

​So, is one set of benefits better than the other? Our answer is no. Both tangible and intangible benefits are important. But only tangible benefits can be used to calculate the financial return of AI investment. This can be looked at from the perspective of additional savings or income generated as a result of AI.

​However, the challenge for many CFOs when it comes to implementing new technological solutions for their companies is clearly defining how success will be measured and quantifying the ROI.

​Since the adoption of AI technologies is not yet widespread but still in the pilot phase we suggest CFOs take a simplified approach to calculating the value of AI projects and follow these steps:

1. Identify a specific problem. Although AI is promising to be a huge game changer for your business, AI is not the answer to all your business problems. Don’t fall into the trap of investing in AI for the sake of investing, or worse, succumb to “herd mentality”. To successfully benefit from AI, first identify a specific problem that may be solved though AI. The AI Identification Worksheet discussed earlier can help you here.

​2. Define the outcomes. What will success look like in your company? What is the result you are targeting, and can this be defined in monetary or percentage values?

​3. Measure the results. After clearly defining the outcomes, the next step is estimating the performance of AI against your baseline measurements or outcomes. The spread between your expected performance and the baseline provides with the expected benefits of the proposed AI solution. Put in place a system to measure the actual results

4. Identify and calculate the costs (investment) incurred in delivering the results. Here you need to consider things like initial investment costs, ongoing support costs and the impact on cash flow.

​5. Calculate the return on investment (ROI). This final step involves calculating the ratio of money gained (or lost) relative to the amount of money invested (the total cost). If the projected ROI meets your hurdle rate, you’ll move ahead with the project. Set up to schedule to review the actual performance vs. the expected results to develop the feedback loop to improve your investment model.

Below is an example of calculating the ROI using the steps above:

1. Identifying a specific problem: ABC Company P2P process is highly manual and incurs annual labor costs of $300,000. During a cost and profitability analysis exercise, Brenda, the company’s CFO established that due to high error rates and rework as a result of these manual processes, the company is incurring additional overhead costs of $100,000 per annum. She remembered that from one of the CFO conferences she attended, the speaker spoke about AI and the technologies potential to drive process efficiencies. She proposes to the Board that the company invests in AI, specifically for improving P2P and test the concept.

​2. Defining the outcomes: After a series of meetings with various functional leaders, stakeholders and consideration of various factors, Brenda presents to the board her findings. By piloting AI for the P2P function, the company stands to achieve annual labor cost reduction of 10% and overhead reduction of 15%. The Board approves the project, expecting savings of $45,000 excluding the potential benefits from higher accuracy and improved vendor relations.

After conducting a thorough market analysis of the suitable AI solutions available, with the support of the Board, Brenda engaged the services of FinancePro, a cloud-based software provider specializing in AI software for the CFO office.

​3. Measuring the results: After conducting a thorough market analysis of the suitable AI solutions available, with the support of the Board, Brenda engaged the services of FinancePro, a cloud-based software provider specializing in AI software for the CFO office. It is now 12 months since the pilot project went live and the Board wants to know if the company managed to achieve the 10% labor cost and 15% overhead cost reduction targets. Brenda compares last years’ costs against current years’ costs and her targets of 10% and 15% cost reductions have been met. In year 2, the company estimated benefits of $60,000.

4. Identify and calculate the costs (investment) incurred in delivering the results: Although the cost reduction targets have been met, Brenda believes that these figures evaluated in isolation are not helpful for evaluating the overall investment. She therefore decides to identify and calculate the total cost ABC Company incurred in meeting these targets. She takes into account all initial costs such as license fees of the new AI software, implementation costs and employee training costs for the full amount of $30,000. She also calculates ongoing costs such as maintenance and support, communications and data storage costs which amounted to $20,000.

​5. Calculate the return on investment (ROI): This is calculated as follows

​ • She uses a cash on cash analysis to determine the 2-year ROI:

​In this example, ROI is calculated by taking the total financial benefits ($105,000) subtracting the total financial costs ($70,000), dividing by the total financial costs then multiplying by 100 to arrive at the ROI (50%). This calculation is over a 2-year period but can be applied on an annual basis as well. We have developed a simple model to help you summarize and compare your AI projects. Use it to:

  1. ​ Analyze and select AI projects,
  2. ​ Get your executive team familiar with the financial benefits of AI and,
  3. ​ As a performance measurement and improvement tool once an AI project has started.

Click here to get your AI ROI Calculation model.

Next Up: Part IV Getting After It: Take the Next Step and Make Your Investment in AI

A Practical Approach to Using Artificial Intelligence for CFOs – Part II

Part II The Benefits of AI and What You Will Need to Make It a Success

If you haven’t had a chance to read Part I – Leveraging AI in the CFO Suite yet, please do so before continuing on.

Potential Benefits of AI for Finance

The potential applications of AI are varied and being considered in virtually all sectors and industries. Today, companies are using AI algorithms to predict start up success, block spam messages and comments on social media, and boost webpage ranking. Lawyers are leveraging the same AI software to speed up legal research, and Financial Advisors have recently been piloting AI to monitor huge data sets and provide data-driven decisions. This handful of uses points to an exciting AI-driven future.

The Finance function is no exception. According to one of the CEO survey findings on the performance of their CFOs published by KPMG, although CEOs are increasingly expecting their CFOs to play an important strategic business partnering role, the gap between CEOs expectations and the actual performance of CFOs is still huge.

CEOs believe, instead of helping them understand and address the business challenges they are facing, CFOs are spending significant time on financial reporting as well as compliance and regulatory issues. In the eyes of the CEOs these activities are more rear-view focused and do little to help them prepare for an uncertain and volatile future.

AI has the potential of helping CFOs close this gap. AI technology can help CFOs automate end-to-end financial processes, make them much more efficient than previously, and spend reduced amounts of time and resources on repetitive and laborious tasks. This in turn helps them spend more time on strategic issues partnering with the business.

Examples where Finance can benefit from AI include:

1. Invoice Processing: Employees spend a significant amount of time on Procure to Pay (P2P). Manually entering invoice data resulting in high and costly error rates. Using an AI powered system, CFOs can significantly simplify and automate these manual processes. Because of the many data points on an invoice, an AI system can “learn” the relationship between the individual elements of an invoice. In the future, based on previous experience and data, the system autonomously processes the invoices and allocates them to the appropriate general ledger accounts. If there are any misallocations which are corrected by an expert, the system learns and improves from such interventions.

2. Bank Reconciliations: The reconciliation of account data and receipts as well as the allocation of banking information can be carried out faster and reliably using AI. The software retrieves both sources of data directly. Independently-learning algorithms match the document information with the transactions in the company’s bank accounts. This renders the bank reconciliation process much more reliable, transparent, and most importantly it can be carried out in real-time. This in turn helps CFOs to evaluate in real time the liquidity position of the business.

3. Budgeting and Forecasting: By using AI, CFOs will be able to improve the accuracy of their company’s forecasts, speed up and automate closing the books with lower compliance and auditing costs. Traditionally, CFOs have relied on financial data housed in ERP systems to drive budgeting and planning processes. This reliance on internal data alone to drive key performance decisions excluded important external data. Thanks to today’s advancements in computing processing powers and speed, CFOs are now able to make use of data sources once deemed inaccessible. AI algorithms are able to aggregate data from multiple data sources, analyze this data very quickly (in real time), identify patterns, calculate the probability and impact on business performance and feed that information into the forecasting model.

New skills, expertise and knowledge required to deliver and operate AI systems

As with any other new technology or system, delivering and operating AI systems requires new skills, expertise and knowledge. New technologies are enabling CFOs to do more with less and create added values for the organizations. Finance and Technology used to be miles apart. Not anymore, the two are now joined at the hip. The CFO has to be tech-savvy and possess a stronger understanding of the new technologies in the market, how easily they can be integrated into the company’s overall IT infrastructure, and their potential to drive business performance.

In addition to having knowledge of the technology landscape, these skills are also a prerequisite:

1. Quantitative: To successfully support effective decision making, CFOs have to make sure that the advice given to business partners is evidence-based and not mere guess work. Having strong analytical capabilities is therefore critical. As data volumes and types continue to grow at exponential rates, making sense of it means the traditional skill set of the Office of Finance has to change. New data analysis capabilities are required; developers, data scientists, data engineers, data architects, data visualization experts, behavioral scientists and cyber security experts working together with traditionally trained Finance professionals.

2. Deep Process Knowledge: Tasks where the desired outcome can easily be described and there is limited need for human judgement are generally easier to automate. Not all Finance processes are candidates for automation. Some processes are higher-value adding requiring judgement or creativity, and are therefore not easily automated. The CFO must be able to differentiate between transaction processing and value-add processes and select the suitable ones for applying AI technology.

3. People Management: Leadership, communication and change management abilities are all essential. Whenever there is talk of AI, the conversation ends up being a debate of Machines versus Humans. There is a common belief that AI has evolved to replace workers. We believe this theory is far-fetched. Implementing AI is also about people and not software alone. Automation is a huge opportunity but it’s also about “augmented intelligence”. In other words, combining human intelligence with technology-enabled insights to make smarter choices in the face of uncertainty and complexity.

The CFO must be able to address any employee fears that might arise, clearly communicate the rationale for adopting AI, and motivate and inspire their team to embrace the change. People are often the differentiator between success and failure. If they don’t buy into the vision of what the company is trying to achieve, the initiative is bound to fail. Also, emotions rise high during such initiatives because of conflicting priorities and as such, it is important for the CFO to manage and resolve such conflicts.

A recent article published by McKinsey in the Harvard Business Review¹ highlights another skill that is important as organizations start working with these new technologies – Data Translator. According to the authors of the article, translators are neither data architects nor data engineers. They’re not even necessarily dedicated analytics professionals, and they don’t possess deep technical expertise in programming or modeling.

Translators draw on their domain knowledge to help business leaders identify and prioritize their business problems, based on which will create the highest value when solved. They then tap into their working knowledge of AI and analytics to convey these business goals to the data professionals who will create the models and solutions. Finally, translators ensure that the solution produces insights that the business can interpret and execute on, and, ultimately, communicates the benefits of these insights to business users to drive adoption.

Thus, as the role of CFOs increasingly evolves into that of a strategic advisor or internal consultant, it is imperative that CFOs develop and improve on these data translation skills. In today’s data-driven era, where data science skills are in high demand, not all of us are cut to be data scientists.

¹Nicolaus Henke, Jordan Levine and Paul McInerney, “You Don’t Have to Be a Data Scientist to Fill This Must-Have Analytics Role?” Harvard Business Review, February 5, 2018

Next Up: Part III Where to Invest in AI, How to Measure the Financial Impact and Select Projects

© 2019 ERPM Insights

Theme by Anders NorénUp ↑