Digital transformation strategy was a trending topic well before the pandemic, but COVID-19 turned it into an imperative overnight.
Organizations such as Zoom and Amazon naturally reaped the benefits of the transition to life under quarantine, but the less obvious winners were almost exclusively the companies with a broad digital footprint — think Etsy, Grubhub, Starbucks, and Pinterest. For these businesses, cloud capabilities made the difference between surviving and thriving when the pandemic struck, and early digital investments put them in a position to capitalize on key technology investments.
There are many technologies that effectively complement digital transformation efforts, but according to research from KPMG, artificial intelligence stands out to both 88% of small business leaders and 80% of those at the helm of larger organizations. And though AI is often thought of as an innovation of the future, it’s far from science fiction. In fact, the benefits of AI for business are well established, and the technology is already firmly entrenched in our daily lives, powering our cars, feeding us entertainment recommendations, making product suggestions, and curating our social media news feeds. In the right setting, AI initiatives are a cost-effective means to propel digital transformation, helping companies collect data, clean it, and mine it for game-changing insights.
Clearing the AIr
AI is often used interchangeably with “machine learning” and “automation,” but there are distinct differences between the three terms. To make sure we’re all on the same page, let’s quickly break down what we mean when we refer to each:
- AI: Artificial intelligence is a broad category describing “smart” technologies that can process new information, react to different situations, and navigate uncertainty. In other words, AI can take on tasks that have historically required human intelligence (hence “artificial” intelligence).
- Automation: If AI is smart — reacting to changes and applying knowledge to new scenarios — automation is simple programming. To perform a task, automated business systems follow a strict set of rules. If those rules are broken or an automated system receives new inputs it doesn’t recognize, exception processing moves the task back to be a manual task.
- Machine learning: Machine learning is a subset of artificial intelligence in which machines take in large quantities of information and process it quickly using algorithms. Unlike the algorithms underpinning automation, machine learning algorithms are also designed to evolve over time in response to certain kinds of inputs — hence the “learning” part. Machine learning in business is typically used to spot insights in vast quantities of data that humans could never sift through.
AI, machine learning, and automation might be three distinct terms, but that’s not to say there can’t be overlap between them. In addition, each of these tools can add value to a digital transformation initiative if implemented in the right place.
AI in Action
AI has exciting potential. And although it’s being touted as a possible solution for all kinds of problems, it’s often easiest to see the benefits of artificial intelligence in existing use cases. In this section, we’ll examine three uses cases in which AI is empowering the switch to digital and generating incredible value for the companies relying on it.
Use Case 1: Supply Chain Verification
Trust Your Supplier is a blockchain network Chainyard built on the IBM Blockchain Platform to help manufacturers combat counterfeit products and build networks of trustworthy suppliers. At its core, TYS offers three valuable capabilities, each powered by a type of artificial intelligence:
- Powerful document and collateral verification: TYS leverages a type of artificial intelligence called natural language processing to go through documents provided by suppliers, identify the presence or absence of key information, and generate a supplier qualification score according to smart rules. The tool performs these mundane tasks with much higher efficiency and accuracy than humans, and it can improve over time as its models learn from processing thousands of documents.
- Improved supplier affinity: TYS can leverage artificial intelligence technology to build intelligent profiles of suppliers based on key criteria that have driven past results. If a supplier provides inaccurate or error-prone documentation, for instance, the supplier profile will be updated to reflect this oversight and the tool can make suggestions about future connections.
- Increased invitation efficiency: Supplier outreach takes valuable time, and it’s only productive when suppliers respond positively and join an ecosystem. By leveraging predictive analytical models, TYS makes outreach more efficient and improves the acceptance rate of supplier invitations.
Use Case 2: Qualifying Loan Applicants
A mortgage is often the most significant investment a person makes in his or her lifetime, which is why Home Lending Pal: Intelligent Mortgage Advisor is designed to help buyers find the right mortgage product for them. By analyzing thousands of data points using machine learning — including existing debt, credit scores, income, and expenses — Home Lending Pal points buyers toward properties they can actually afford and suggests lenders that will be willing to loan them the money they need. Home Lending Pal is also improving loan access for customers with no credit history who wouldn’t otherwise qualify for a loan.
Although AI is certainly helping connect buyers with mortgages, it’s also being used to help banks predict how likely customers are to repay their personal loans. Upstart is one such tool, and the lending platform works with banks to augment limited credit scores (or replace them if credit scores aren’t available) using factors such as education and employment status. With AI predicting repayment, banks are less likely to lend to customers that will default on their payments.
What does adoption look like across the industry, though? A 2018 survey by Fannie Mae found that only one-third of mortgage lenders were utilizing AI, and about half of them were merely experimenting with the technology. That number is on the rise, however, and the same survey found that just 2% of lenders wouldn’t be willing to use the technology at all.
Lastly, the mortgage and lending process is just one potential application of artificial intelligence, and McKinsey’s Global AI Survey found that almost 60% of respondents in the financial services sector have adopted at least one AI capability. Robotic process automation is the most common, followed by chatbots or virtual assistants for customer service teams and machine learning tools to spot fraud and augment human underwriting teams. Although many financial-services organizations are still adopting AI in response to a specific problem, a growing number are seeking to implement it more broadly throughout their organizations.
Use Case 3: Digital Workers
A digital worker is a kind of software solution powered by various applications of artificial intelligence, ranging from natural language processing to machine learning to computer vision. Instead of supplanting human workers, digital workers perform tasks alongside them with speed, efficiency, and even advanced decision-making capabilities.
Digital workers have the capacity to transform the workforce in two key ways:
- Operating capacity: Digital workers never go on vacation or even sleep, and they’re immune to human concerns such as boredom and burnout. The “always-on” software can accomplish huge quantities of work, allowing companies to serve more customers, improve operating margins, and price products and services more competitively within the marketplace.
- Worker experience improvements: Digital workers can take on the more mundane tasks necessary in an organization and give humans the freedom to focus on more strategic, higher-value work — but that’s certainly not all they can do. Because they’re powered by artificial intelligence technology, digital workers can offer insights and suggestions that actually improve the decision-making abilities of humans.
Digital workers are already having an impact in a number of industries, and their influence will only grow. According to research from IDC, digital workers will contribute 50% more to the global workforce from 2019 to the end of this year, with a 28% increase in instances of technology evaluating information and an 18% growth in reasoning and decision-making implementations.
Building AI and Machine Learning Into Your Digital Transformation
Artificial intelligence as an idea has existed for decades, but the amount of high-quality data available and steady advancements in processing power have made AI a burgeoning field full of exciting possibilities.
1. Educate yourself.
Before you can get an accurate picture of AI’s potential impact on your organization, you need to understand the different types of cognitive computing, how they’re deployed, and
how they’re applicable to your business. Robotic process automation, for example, involves the automation of both digital and physical tasks; it’s the least expensive option and can offer the quickest payback period of any artificial intelligence technology solution. On the other hand, cognitive insight, which uses machine learning to detect patterns in vast volumes of data, can offer incredibly valuable insights — but the payoff isn’t guaranteed.
2. Look for inspiration in your industry.
Whether it’s through a robotic investing advisor, a drug discovery tool, or a customer service chatbot, AI implementations are often specific to industries — and you can skip a lengthy discovery process by looking at the benefits of AI for businesses in your sector. Examine how your competitors are applying machine learning to business problems for inspiration, and look for standout examples of business process automation tools and other automated business systems.
3. Address pain points with the biggest impact.
With each passing day, it seems there are fewer limits to what AI can accomplish, but that doesn’t mean your first implementation should address the most complex issues in your organization. Look for obvious pain points where the technology can unlock the biggest benefit; this is usually done by eliminating an existing bottleneck or automating a manual process to allow your organization to scale. For example, if your business is ready to serve new customers but can’t seem to find sufficient suitable prospects, a tool to comb lead databases and support your sales team might be the best investment. If you already have plenty of customers but satisfaction rates are suffering, a chatbot can help address many of the most common queries and ease the burden on your service personnel.
4. Launch pilot projects.
Cognitive applications should always start with a pilot project that allows you to learn about the technology, understand how it will be integrated into your environment, and evaluate the capabilities of your staff. Develop a center of excellence around new technologies to help your organization scale a solution across multiple departments. If you notice that you’re missing certain capabilities internally, you’ll need to bolster your team by relying on third-party vendors.
Identifying a Promising Partner
AI implementations are complex undertakings, which is why it’s common for companies to partner with vendors who can bring advanced skillsets and a wealth of experience. Not all vendors are created equal, however, and because a capable partner can make or break an implementation, it’s important to keep a few things in mind when choosing a third-party provider:
1. Viability.
You don’t want a new vendor that’s going to use your company as a learning opportunity. Look for a partner that has been around for more than five years and has a track record of success. These companies will have the fiscal security and corporate maturity to be dependable not just now, but also for years down the road.
2. Support.
An AI implementation is a journey and not a destination, so don’t expect a project to operate on autopilot once the implementation process is complete. Look for a vendor you can return to for help with ongoing needs, and one that can fit you in for future projects as well.
3. Flexibility.
When choosing the right AI solution, the decision should be based entirely on your needs and not on the preferences of a potential partner. If vendors only advocate for the specific flavor of AI they specialize in, it’s safe to assume they’re more interested in their own success than in yours. Along those lines, look for a vendor that embraces open source over pushing the proprietary technologies it sells.
4. Adaptability.
Your organization’s goals should be at the forefront of any vendor’s work. Although vendors should have their own proven processes, they should be willing to adapt their techniques and procedures to your team and your organization’s style, culture, and mission. If they’re not willing to be flexible, the partnership is unlikely to be a productive one.
Incorporating AI and machine learning in business should be a part of any digital transformation strategy. The benefits of AI for business are well documented, and unlocking these benefits in your own organization is simply a matter of understanding the technology, identifying its most promising applications, and assembling the capabilities necessary for a successful implementation.
For more information about how AI in digital transformation could impact your organization, contact Chainyard to consult on how we can work together.
