Across industries throughout Asia Pacific, digital transformation is a near-universal goal, but data is what’s really setting the terms — and accelerating the pace — of digitalization.
The premise is simple: the better the curation of data, the more you can achieve with it. Organizations are collecting unprecedented amounts of data, and they need a means of efficiently storing, accessing, and analyzing all that data to deliver business value.
Where a typical enterprise once took in structured data from applications and stored it away for later success, that same business will now need to handle many new varieties of unstructured data — think images, video feeds, and audio recordings, to name just a few.
By 2025, IDC predicts the current storm of data will exceed 175 zettabytes a year globally, with 80% of enterprise data expected to be unstructured. As data growth continues to spread across industries — including healthcare, manufacturing, retail, financial services, public sector, media, and entertainment — every enterprise faces an urgent challenge in finding a way to overcome the unstructured data chasm to establish market advantage while adapting to the new normal.
DigiconAsia gleaned some insights from Asheesh Mehra, Co-Founder and Group CEO, AntWorks:
What are the biggest challenges businesses are facing due to the growth of unstructured data?
Asheesh: A major challenge that businesses have been faced with is having insufficient resources and tools to deal with the sheer volume of data each enterprise can now generate – 80% of which comprises of unstructured data that needs to be processed and analysed before it can add any value to the business. Couple this exponential growth of unstructured data alongside the impact of COVID-19 and today we are seeing businesses forced to accelerate their digitalization and automation journeys or risk being left behind.
We’ve witnessed a significant uptake of enterprises seeking help with firstly identifying the processes they should automate, with many expressing interests in Process Discovery, a Machine Learning based technology that records all possible variations of a process and makes recommendations for how to automate. While it has become a necessity for business to adopt some form of automation to boost their operational resilience, leaders must be able to identify the most effective opportunities for automation at the outset. Without such knowledge, it can inevitably lead to failed projects and disappointing outcomes.
During this unprecedented time, it has become even more imperative to get the “data piece” of the automation puzzle right at the outset. The advent of COVID-19 has further highlighted the failings of traditional technologies such as Optical Character Recognition (OCR), and the time has now come for businesses to look to solve for their unstructured data challenges to build a stronger, more robust and future-proof business post the pandemic
What are the potential opportunities hidden within unstructured data, especially for organizations in Asia Pacific?
Asheesh: How good an AI application is depends on the data it uses, and its capability to deal with the different types of data it’s presented with. The real problem that businesses in APAC have to get their mindset around is the exponential growth in unstructured data and the fact that its only increasing year-over-year. In short, it’s a big problem already which is only going to get bigger. Tackling it now really is the only course of action that makes competitive sense.
Every day documents are being created and stored by every part of every business — with more and more data piling up each day. Needless to say, organizations with operations that are paper-intensive stand to benefit the most from automation, however it’s a problem for every business irrespective of sector or size.
Take, for example, all the paperwork that goes into applying for a mortgage, a new car loan, or insurance claims. It’s tedious work for businesses to manually process stacks of documents and impedes organisational efficiency, scalability, customer service, and achieving competitive edge in a digital world. Yet the “paper trail” is not going away, as it remains crucial for establishing auditability, accountability, continuity and privacy.
To manage the influx of documents, an increasing number of companies across industries have begun to embrace intelligent automation solutions – of which some may have chosen to adopt OCR. While OCR can deliver some benefits by automating structured data, businesses will find themselves hindered by the tool’s inability to read unstructured data and digitise documents that contain images and inferred data, printed or handwritten text, or simply image data such as notary stamps or signature verifications.
How can AI and automation aid digital transformation and help resolve the unstructured data challenge?
Asheesh: Using a cognitive-based data digitisation platform such as Cognitive Machine Reading (CMR), that incorporates fractal-based Machine Learning and understands all data types, organisations can access greater accuracy and higher processing speeds – empowering them to confidently digitise all of their data, and in turn, overcome the unstructured data challenge.
John Hancock Manulife, a global insurer, used to manually process the large volume of policy management documents it received. Many of those documents held vast amounts of unstructured data, especially handwritten text. Since adopting CMR, they have been able to decrease their turnaround times and achieve up to 92% increase in accuracy for handwritten recognition.
With the right intelligent automation technologies, companies can improve efficiency, retain customers and increase visibility. This is now a time where it has never been more important for companies to leverage their data, not only to make employees and processes more efficient but also to maintain operational resiliency and continue delivering services and customer satisfaction in a post-COVID world.
What are some key considerations to ensure success when adopting AI and automation tools and solutions?
There are three key considerations to keep in mind when sourcing for an AI and automation tool and they are: intuitiveness, speed of discovery and business-user empowerment.
Businesses must keep in mind that adopting any AI and automation solution will require some degree of training, however minimal. To ensure quick deployment, minimise implementation cost and most importantly accelerate time-to-value, they need to look for tools that are simple and intuitive enough for business end-users. This is absolutely critical as additional or lengthy training involving other departments will significantly affect the product delivery timeline and incur added costs.
The solution’s speed of discovery, or in layman’s term – its performance, is also a factor that business leaders will look out for and rightfully so. Some products take days, weeks or even months to collect, analyse and review data that is usable for automation. Hence, leaders would need to exercise judgement in selecting a tool with sufficient speed or risk delays.
Leaders should also consider the level of control they have over the solution. For instance, our ANTstein Process Discovery provides businesses with extensive and flexible user control over ‘what to automate’, enhancing business agility and adaptability. This adds on a layer of adaptability which is beneficial for the business should there be a need for it to re-adjust its objectives due to the changes taking place in the industry.
