Artificial intelligence has gone mainstream. Companies across industries have conducted successful proof-of-concepts and even successfully deployed AI in production. Some businesses have even implemented their AI and machine learning strategies, and various projects have proliferated across the business, while also being outfitted with best practices and pipelines. Today, companies at the forefront of the AI maturity curve are using AI at scale.
The overall maturity of how AI is deployed in the enterprise is changing how companies view the strategic value of AI — and where they want to reap the benefits of AI. Here are ten strategic trends in enterprise AI that industry experts see today.
1. Artificial intelligence begins to play a practical role
In the early days of AI, projects were driven entirely by data scientists. Data scientists have data and algorithms and are free to find ways to use new tools to solve business problems. Sometimes, they can successfully solve the problem. Today, this situation has changed a lot.
Business leaders have learned from successful projects and learned more about what AI can do for them. As a result, companies are now less willing to invest in proof-of-concepts whose business value is unclear, a trend that shows the business sector is increasingly the dominant player in AI applications.
Alex Singla, global head of QuantumBlack, a consulting firm owned by McKinsey & Company, said: “When I see companies doing well with AI, it’s all business-driven. AI And IT can help a business solve a problem, but it’s not the technology that comes up with the solution.” It’s the business that takes the lead and says, ‘I’m part of the solution, I believe in this solution, this is the right solution. ’”
For example, Honeywell’s chief digital technology officer, Sheila Jordan, said AI is being used in her company’s internal operations and embedded in customer-facing products and services.
“We’re very connected to the business,” she said. “We are driven by value. That will be customer-facing value. Intrinsic value.”
2. Artificial intelligence penetrates the entire enterprise
When Jordan came to Honeywell two years ago, her first big project was to implement a data warehouse strategy to bring together all transactional data from all sources.
“Every function, every business unit has a digital agenda,” she said. Honeywell, for example, has digitized all of its contracts. That means there are more than 100,000 contracts in total, she said. She noted that these contracts provide companies with a wealth of data that can be used to help build AI solutions for almost any functional area.
For example, all of Honeywell’s contracts can now be automatically reviewed for areas affected by inflation or pricing issues, thanks to artificial intelligence, Jordan said. “It’s impossible for anyone to scrutinize 100,000 contracts.”
Likewise, with complete inventory data, Honeywell is now able to understand which inventory is scrap and which is reusable, allowing it to make informed decisions about managing raw materials more efficiently, Jordan said.
“We’re seeing AI appear in every functional area,” she said. “Finance, legal, engineering, supply chain, and of course IT.”
3. Rapidly advance automation with artificial intelligence
This is Honeywell’s third year of aggressively advancing automation projects. If there is repetitive work, companies will try to automate it. “We might have 100 projects this year,” Jordan said. “These projects will allow us to automate certain jobs at companies around the world.”
Honeywell is working to make these automated processes smarter, she added. “We’re going to put more artificial intelligence in these automated robots,” she said. “That means automated robots are getting smarter.”
Another company starting with basic, rules-based automation is Booz Allen Hamilton. “Now we’re integrating AI and machine learning into these automated processes to make them applicable to a wider range of business work,” said Justin Neroda, vice president of AI at Booz Allen.
People start with the simplest automation, he said. “Then they ask themselves, ‘What other jobs can I automate?’ And then they find out that this requires the use of artificial intelligence and machine learning.”
He said AI-driven automation could help companies deal with staffing shortages or handle large volumes of work. “Or half the jobs could be automated, and then people could do the hardest part of it.”
4. Use artificial intelligence for greater benefits
There is also an important change management element for implementing AI at scale, says McKinsey’s Singhella. He said this requires understanding how people will use AI, not from technologists working alone, but from a combination of technologists, disciplines and business experts.
“If I had to use artificial intelligence, I would tell them to learn about three different areas of artificial intelligence that are very unlikely to use artificial intelligence,” he said. “But the more AI is automatically integrated into the workflow, the more likely we are to succeed. The less I change other people’s behavior, the more likely I am to be accepted.”
5. AI strategy needs a collective shift
After a successful initial proof-of-concept, companies often establish AI centers of excellence to implement the technology and develop talent, expertise, and best practices. But when a company reaches critical mass, it becomes necessary to break up some centers of excellence and integrate AI technologies, and bring experts directly into the business units that need them most.
“For companies that are less mature, there is value in having a center of excellence that takes in talent and enables learning across the organization,” said McKinsey’s Singhella. “Without a center of excellence, companies often don’t have a center of excellence. Capability scales up. Talented people want to be around other like-minded people. And less experienced people can benefit from a center of excellence, where they can grow or learn.”
Breaking up a COE prematurely erodes its influence and reduces a company’s ability to iterate and replicate successful projects across multiple business areas.
“But in the longer term, as you reach a certain level of maturity and scale, the benefit of having a technologist with deep AI expertise and domain expertise is to enable real business success,” he said. “But only if you have a certain scale.”
Business problems are scattered everywhere, says Amol Ajgaonkar, a distinguished engineer at Insight.
“Business problems aren’t going to be centralized in one place, so you can’t count on centralized AI deployment,” he said. “These deployments also have to be spread out. But you do need to have a centralized AI strategy that impacts a business.”
Or, he added, an AI strategy that can impact multiple businesses, such as revenue, cost savings or market positioning.
Like many other companies, Booz Allenne started with a core AI team. “But over the past year, we’ve really been expanding that team,” said Justin Neroda, vice president of AI at Booz Allen & Co. “We got some small teams through that company with AI experts. But before you can expand the team, you have to get to a critical mass or it will fall apart.”
“This is what we’re seeing in our own company and in the clients we work with,” he added.
6. AI triggers business process transformation
When businesses first start using AI, they typically look for a single step in a business process where AI can make a difference. “You can break a business process into pieces, digitize each piece, and use artificial intelligence to increase efficiency,” said Sanjay Srivastava, chief digital officer at Genpact. “But at the end of the day, the business process itself is the same. It’s Every part got better, faster, cheaper – but the process itself didn’t change.”
But he said AI also has the potential to fundamentally change business processes. For example, Genpact does a lot of account processing for its clients.
“When we apply artificial intelligence to processing invoices, we can know which invoices are going to be disputed,” he said. “We can find out which part of the portfolio has the highest risk.”
With the predictive power of artificial intelligence, the entire process can be restructured, he said. “When you apply AI, you can think about the end-to-end value chain and you can completely restructure it.”
7. Machine Learning Operations (MLOps) become a reality
According to a McKinsey & Company report released in late 2021, one difference between companies getting the most profit growth from artificial intelligence is whether it uses machine learning operations.
This is the next big thing in artificial intelligence, says Carmen Fontana, IEEE member and head of the cloud and emerging technologies business at Augment Therapy, a pediatric physical therapy technology company. . Fontana was previously the Cloud and Emerging Technologies practice leader at Centric Consulting.
Our goal, she says, is to translate machine learning from theory to practice. “Two or three years ago, it was an emerging field and people thought they had to do it,” she said. “But we don’t see many applications in practice.” Today, however, she sees some mature tools and methodologies that allow businesses to become more rigorous in training, deploying and monitoring AI models.
“This goes a long way towards institutionalizing artificial intelligence and machine learning techniques,” she said. “I’ve seen all of this with our customers. The market has changed dramatically.”
8. Enterprises laying artificial intelligence pipelines
We’re currently working on about 150 different AI projects with clients, says Booz Allenne’s Neroda. But over the past year, our company has begun to move away from this one-off model.
“For the past year and a half, we’ve been investing in modular capabilities and end-to-end pipelines,” he said.
Success in AI requires more than just a working model. He said that as the data changes and the model continues to improve, a complete process is needed to maintain the model.
“The biggest challenge is tying all the tools together,” he said. “We’ve been working hard to standardize it and build some reusable components to use across different projects.”
9. Organizations want to build trust in AI
As employees and executives become more familiar with AI, they increasingly trust AI to make critical business decisions — even when those decisions sometimes go against human intuition.
Michael Feindt, a strategic advisor and founder of Blue Yonder, recently worked with a major UK food retailer that was struggling with supply chain woes related to the pandemic. When the company uses manual processes to manage its supply chain, there are a lot of empty shelves, he said. In addition, there is a shortage of people who are knowledgeable, capable and willing to do the work.
Automated AI systems can reduce costs and improve performance. However, when the pandemic hit, people wanted to turn off automated systems. “But then they found that the automatic system adapts much faster than humans,” he said.
So instead of shutting down these automated systems, the company expanded them to stores and distribution centers. The result is less empty shelves and less food waste thrown away. Additionally, store managers can stop spending two hours a day fine-tuning their orders and spend more time improving customer satisfaction.
Feint said there are other ways to build trust in AI. “Some people are picky and don’t believe AI can make the same good decisions as them based on their years of experience,” he said. Adding some interpretable features will help alleviate these problems. Explainable AI is when the system can explain to a human user what factors led it to make that decision.
10. New business models may emerge
In some areas, artificial intelligence is creating unprecedented opportunities. Self-driving cars, for example, have the potential to transform society and create entirely new types of business. But AI-driven business transformation can also happen on a smaller scale.
For example, banks that require human review are unable to provide small loans. The cost of researching and processing these loans will be higher than the interest income the bank can earn. But if AI is used for evaluation and processing, banks can offer small loans, allowing them to serve a whole new group of customers without charging exorbitant interest rates.
“These use cases are still not that common,” said Jai Das, president and partner at Sapphire Ventures. “These use cases are fundamentally changing the way we work, and businesses aren’t changing that fast.”
That trend will start to turn once AI and machine learning technologies become tools used by every knowledge worker in a company, he said.
“We’re not there yet. It’s probably another five years before people use artificial intelligence and machine learning to do their jobs.”
GIPHY App Key not set. Please check settings