Automation is like a big tent, covering a diverse set of technologies. On one end are the simplest of simple automation tools, like the free If This, Then That (IFTTT) applet. It automates the gaps between consumer applications, such as triggering a photo to be sent to a file folder after a person presses a social media Like button. On the other end are the smart contracts made possible by blockchain. And in between are enterprise-wide applications such as ERP, which incorporate automation and AI capabilities to extend their processes.
PwC has a slightly narrower focus on scalable and currently deployable automation technologies, from no-code tools for simple automation to advances in AI that add intelligence to automation.
Intelligent data capture, combined with simple automation, enabled a global bank to vet KYC documentation in a matter of hours, rather than days or weeks.
A new class of tools for business users can now provide the gateway to automation, but those tools alone won’t make breakthrough gains possible. Selecting the right combination of additional technologies is how intelligence is added to automation. For example, a global bank wanted to automate its know your customer (KYC) processes for global clients, a long process of requesting information, checking against databases in different jurisdictions and then making a judgment. Simple automation eased some of the burden, but the real gains came when the bank added machine vision and natural language processing to “read” documents completed by the customers. That intelligent data capture, combined with simple automation, enabled the bank to vet KYC documentation in a matter of hours, rather than days or weeks.
Natural language processing is an entry point on the journey to AI. Indeed, many companies are still selecting AI use cases and looking for practical ways to apply the technology. Automation may be the answer they’re looking for.
Intelligent automation quietly incorporates an ecosystem of technologies. Many companies, for example, use customer service chatbots on their websites. Such chatbots often start with rule-based automation and natural language processing but are later fortified with additional cognitive understandings of sentiment. Similarly, when insurance companies send drones to recognize the damage to rooftops after a hurricane, they can optimize machine learning image recognition models as an essential step to quick claims adjustment and payout.
Many companies are still selecting AI use cases and looking for practical ways to apply the technology. Automation may be the answer they’re looking for.
While today’s automation toolbox provides plenty of choices for getting started and realizing gains relatively quickly, don’t jump to the technology. Both business and IT leaders should talk about process challenges, what changes people can expect and how you’ll govern projects.
Combining simple and intelligent automation technologies creates a multiplier effect. Build upon RPA efforts, for example, with more advanced solutions that can automatically learn and optimize processes.
If there are any doubts about AI’s practicality in business today, put them to rest by applying machine learning and natural language processing to solve process challenges.