Intelligent Automation (IA) is a newer concept among enterprises that promises to completely reimagine how business is done. That’s a big statement, but one which thought leaders and decision makers around the world are increasingly taking to heart. To truly understand the potential of IA and what it means for the enterprise, we need to break the term down to its fundamentals.
Automation: Automation is the use of machines to execute tasks that previously required human intermediation — and typically, those tasks are completed far more efficiently than a human ever could handle alone. Physical automation has a history reaching back centuries when machines began executing or augmenting tasks that previously relied on humans or animals. Next, computers came along and began executing computational and transactional tasks error-free and at the speed of electrons, which eventually lead to the rise of Robotic Process Automation (RPA). In more recent years, Artificial Intelligence (AI) has added a dimension of cognitive functionality to automation, thus allowing digital systems to replicate intelligent processes that previously could only be handled by a human brain (e.g. decision making, visual recognition, learning or natural language recognition).
Intelligent: AI is an especially important development because it empowers machines to execute tasks they weren’t specifically programmed to anticipate; it’s why a driverless car can navigate a route it’s never driven before, or why a voice assistant can process a sentence it’s never previously encountered. AI enables computer systems to not only handle complex operations, but improve at them over time through self-learning and experience — all with limited-to-no human intermediation.
Combine these two principles together, and you can achieve Intelligent Automation; IA is what happens when AI is used to automate or augment the creation of new automations, and it is opening the door to a number of intriguing possibilities for the enterprise.
The most direct result of IA is a dramatic increase in the number of automations. Just as other forms of automations increase output, automating the creation of automations naturally leads to an increase in automation creation.
In a previous technological era, it could take days, weeks, or even months to write a new automation depending on its complexity. IA can greatly accelerate this process by independently creating new automations from scratch based on high-level commands (e.g. an HR manager can instruct an HR system with IA to send a personalized email to all employees informing them how many days off they have left at the end of each quarter), or it can augment the process by simply automating more time-consuming tasks required for automation to be effective. For example, 1RPA automatically structures unstructured data and intelligently identifies and “understands” web pages, emails, programs or any unstructured text it encounters. This automation allows engineers to slash the time needed to create an automation, thus allowing them to be that much more productive.
So, whether IA is completely automating the automation-writing process, or merely augmenting it, it inevitably results in the production of more automations, which can dramatically alter the nature of a business environment depending on how automations are utilized.
Intelligent functionality not only leads to more automations, but also supports the production of higher-quality ones. For example, IA can power automated A/B testing solutions, which have been used to great effect in the marketing, content and editorial fields. Automated testing allows digital systems to independently test numerous versions of a discreet input or element (say, subject lines on an email) through randomized micro-experiments, and then analyze the results to determine which version was more effective based on specified metrics (for example, open rates on particular email campaigns). Furthermore, Machine Learning (ML), a sub-field of AI, enables the system to discern lessons from its micro-experiments to utilize in the future. For example, it may learn that using all capital letters in an email subject line results in an average of 20% fewer opened emails, and then use that element moving forward.
While this sort of IA-guided automated testing has been used in areas such as creating acutely viral content, the same principle of automated experimentation and analytics can be used to create impactful automations, which thanks to ML will only improve over time.
Opening the Process
When IA automates many of the arduous processes of writing new automations, the process can be opened to users who don’t have technology-related degrees or extensive experience in coding. Similar to how the first GUIs opened computing to the general public by simplifying complex computing processes for human sensibilities, IA can lower the bar for crafting new automations to more users.
1RPA contains an intuitive User Interface (UI) by which users press a virtual “record” button and perform a task while the system observes their actions and translates it into a new automation (you can watch that process in action here). To lower barriers to entry even further, 1RPA comes packaged with the digital colleague Amelia, so users can have a conversation with the system in a human-like manner. Specifically, users can simply inform Amelia (via a chat interface) the task that they would like to accomplish, and she can lead them step-by-step through the creation process.
IA has the potential to accelerate automation adoption, while at the same time essentially making those automations smarter over time. As the IA paradigm takes hold in various industries, a sharp divide will quickly develop between enterprises which have added intelligence into their automation mix, and those who haven’t. What side will you be on?