Across the globe, artificial intelligence (AI) continues to have a fundamental influence on business due to the pandemic induced disruption and growing use cases.
Various studies show that AI adoption is nearing a tipping point and would shortly become a ubiquitous technology. Almost one-third of IT professionals in a global survey by IBM Institute for Business Value (IBV) said their company was using AI, even as half of them revealed their companies were exploring the technology.
In India, over half of Indian IT professionals reported that their companies had accelerated the roll out of AI. As AI adoption rises, what is the roadmap that companies should consider, particularly those who are either evaluating adoption or have not yet achieved maturity in deployment?
Before unravelling the game plan, it is crucial to understand the factors driving AI adoption by companies from varied sectors.
Also Read: Three ways AI can help transform businesses
What is driving sustained AI adoption?
Companies are adopting AI to improve customer engagement and service through natural language processing (NLP) tools. Considering they collate vast amounts of unstructured language data from customer emails, reviews, and social media posts, they are looking to identify insights from such data to enhance customer experience and prevent problems even before they happen.
With NLP, they can differentiate themselves from the competition, ring fence their reputation and create more cross and up-selling opportunities.
Moreover, the rise of intelligent automation is also pushing AI adoption. Today, innovation-led companies are infusing automation with AI for enterprise-wide deployment so computers can make complex and mission-critical decisions themselves with the limited intervention of human agents. In the COVID-19 era, companies need real-time insights so their operations can be dynamic, responsive, and interconnected to their work flows and the ecosystem.
Intelligent automation can catalyse innovation, enhance customer experience and efficiencies and save costs by transforming business processes into intelligent workflows.
Over 50 per cent of the respondents of the survey cited earlier said they used automation to interact better with customers during the COVID-19 pandemic.
Companies are also using AI-infused automation to optimize business processes - for instance, with intelligent document management tools, financial institutions are classifying documents and quickly extracting data for processing loans based on documents submitted by their customers.
By deploying robotic process automation bots with low-code editors, companies in various sectors are automating repetitive activities, significantly reducing copy-and-paste and data-entry errors.
Besides, they are also simplifying their IT operations and resolving IT issues faster with automation. According to the IBV study, about 62 per cent of Indian IT professionals cited tremendous efficiencies for processes and tasks for considering using automation. Now that the factors for adoption have been outlined, here is the roadmap companies can follow in their AI journey.
Recommended roadmap for AI adoption
Companies that are still evaluating or piloting AI adoption can take several measures to achieve scale rapidly. The first step is to use the minimum viable product approach to identify and prioritise projects based on complexity, business impact and risks.
Not every pilot may succeed and reach production, but this approach is recommended instead of a proof-of-concept method considering AI is already a proven technology.
Besides, companies should not wait for data governance projects to roll out AI as it can happen in parallel; all they need is to understand what data they have, where it resides, and who manages it to increase confidence in the output.
Most importantly, they should involve their business leaders by embedding them into AI projects following adequate training in AI skills, if necessary.
The key is to focus AI projects to support the larger strategic agenda of the company and not implement them in isolation.
Companies that have already deployed or operationalised AI projects too can chart a roadmap to reap the full benefits the technology affords.
As a first step, they should establish an AI play book replete with checklists and engineering principles built upon successes, failures, and key performance indicators.
It is vital to create an architecture and team structure that operates at the intersection of design and data centres. It is also critical that companies evaluate and iterate their AI models while in production to achieve scale.
Besides, they must understand that if their AI models are not repeatable then they are not reliable.
It is where documentation plays a significant role. Notwithstanding the maturity levels of AI deployment, companies need to prioritise building trust amongst stakeholders such as customers, regulators, and governments to achieve scale.
They must be able to explain to end-users why and how their AI model makes certain decisions. Together with the 'build for performance' approach, they must focus on the fundamentals of fairness, explainability, robustness, transparency, and privacy to foster trust in their AI systems and business outcomes across the entire life cycle.
To conclude, the sustained focus on AI deployment by companies to solve a spectrum of business challenges will make the technology commonplace. Besides, AI adoption will get an impetus from hybrid cloud migration, the maturing of the digital transformation agenda and the desire to become fully virtual through automation.
But as companies adopt AI, they must focus on the right strategies to maximise their returns and leverage the technology to the fullest.
(The author is Vice President, Technology, IBM Technology Sales, India/South Asia.)
Copyright©2022 Living Media India Limited. For reprint rights: Syndications Today