Big waves of business process reengineering have historically been few and far between. One reason has been the length of time it takes humans to analyze trends, conceive new ways of operating, and then integrate the process changes into software code. Machine learning (ML) is knocking down that barrier.
ML algorithms’ ability to adapt on their own, requiring no explicit programming to learn from changing conditions, lets them continually hone their automated problem-solving, predictions, and insights. The more good data they’re fed, the smarter they become. As such, ML is enabling organizations to rethink existing processes at unprecedented scale and speed, improving efficiencies and spurring innovation at a pace that once would have been unimaginable.
Innovation and automation across industries
Machine learning is addressing all manner of use cases for process automation. Here are three that demonstrate the breadth of ML’s impact on an organization’s day-to-day activities:
- Quality control: Swedish food manufacturer Dafgards set up a proof of concept (POC) using Lookout for Vision, a computer vision technology, to spot defects, anomalies, and mismatches in its pizza production lines. The technology allows Dafgards to automate pizza production line inspections to verify that all pies have the right amount and distribution of ingredients and to detect anomalies, such as the wrong topping. Used this way, ML makes Dafgards’ processes less error-prone while helping keep its customers happy. After the success of the POC Dafgards is looking to roll out the services on multiple production lines.
- Document processing: Automating data extraction can reduce time and costs required to manually capture information from many types of documents, including invoices, patient intake forms, loan applications, contracts. ML can be applied to extract text from millions of documents, in context, quickly and economically. For example, consumer finance company Dealnet Capital has reduced the amount of time it spends reviewing loan applications and other documents by up to 80% using cloud-based ML services.
- Customer service: In the modern contact center, ML empowers agents to understand customers’ history and make recommendations accordingly, helping to reduce resolution times and improve customer satisfaction. Contact centers can also leverage ML-powered chatbots and voicebots to quickly assist customers. The need for rapid access to up-to-date information was particularly pressing for government agencies during the COVID-19 crisis; public sector organizations had to quickly develop new ways to share information or improve on existing systems to support a surge of website visitors during the pandemic. The state of Rhode Island’s Department of Labor and Training (DLT) implemented Amazon Connect, an omnichannel cloud contact center, to replace outdated interactive voice response (IVR) and interactive web response (IWR) systems, which in the early days of the pandemic were struggling to support 10 times the typical volume of unemployment insurance applications. Amazon Connect, with built-in AI and ML capabilities, enabled the agency to increase capacity from 74 concurrent calls to up to 1,000 concurrent calls per minute.
Whether ML is used in these ways or countless other applications, one thing is certain: Machine learning is creating a new model for improving the processes and workflows that support any enterprise.
Learn more about ways to reinvent your business with data.