The Different Stages of Automation: What You Need to Know

By Juan Martinez, Senior Writer
January 31, 2019 • 5 minute read

Don’t assume all automation tools are created equal. Find a solution backed by AI and cognitive intelligence, like IPsoft’s AIOps (formerly 1Desk). Read this article to see where you fall on the automation lifecycle journey.

As your company updates its applications and business systems, it’s crucial that you investigate the power of automation. Whether you’re automating repetitive back office tasks, or deploying virtual engineers to perform complex end-to-end IT operations, automation can make your business more productive and cost-efficient.

There are a variety of automation solutions available in the market today; they are not created equal and each comes with various capabilities and features.  Enterprises need to be educated about the different levels of automation that exist. Before choosing an automation vendor, you need to determine where your business is on the automation lifecycle and, perhaps more importantly, where you ultimately envision your business to be.

Simple Tasks Triggered by Simple Tasks

Let’s start with the most basic level of automation: Simple actions that are activated by simple actions. For example, when you add an item to a shopping cart on a website but you don’t make a purchase, retailers send you follow-up emails to remind you that your item is still sitting in the cart. This isn’t done manually, but instead via an automation that gets triggered after an item sits in a cart for a pre-determined amount of time. This level of automation has existed for more than a decade and can be applied to all lines of business. IT teams can create simple triggers that activate changes to a network when traffic thresholds are met or exceeded. Human Resources teams can automatically send emails to managers when workers request vacation days. There are even companies that can help you build simple automations that span multiple tools, so when an action is taken in system A, it triggers an action in system B, which then triggers an action in system C.

Odds are you zoomed past this stage of automation years ago; if you haven’t, you should consider building a technology team with deep knowledge of enterprise software.

Chatbots and Self-Service Automation

Most chatbots and conversational interface-based tools claim to offer automation. Some, however, are actually just decision-tree systems that respond to keywords. Think of automated voice messaging systems: When the system asks if you want to speak in English, you say, “Yes,” and the system is prompted to take you on a different path than it would have if you said, “No,” and opted for Spanish.

Chatbots and self-service-based customer service/service desk tools function the same way. As you type/text back and forth with a virtual assistant, you are led to believe that there is advanced automation and intelligence on the other end of the conversation. In fact, some chat-based tools are programmed to follow basic decision trees with pre-determined pathways. So while there is a rudimentary kind of automation involved, these tools are optimal for only very simple tasks, such as consumers contacting a business about store hours or return guidelines. These are static, generic answers that are easily identifiable with keywords, so a simple chatbot can accurately provide responses.

RPA Without Intelligence

Robotic Process Automation (RPA) is another form of automation that doesn’t require AI. In most RPA scenarios, an IT team performs a task on a computer, and the software records and repeats it on its own for identical tasks in the future. Unfortunately, RPA without intelligence can require a team of coders days or weeks of labor to build basic automations, and then if one change occurs to a business process, the automation has to be completely redone.

Additionally, unintelligent RPA requires manual labor to make sense of any data that doesn’t follow a particular structure. If someone sends an email and the information contained within it needs to be processed to fulfill an automation, someone needs to put that data into the proper sequences so it can be properly logged. This is both time-consuming and antithetical to the point of using RPA.

Nonetheless, if you have tasks that are repetitive and time-consuming, and that don’t require much system intelligence, RPA can be incredibly helpful. For example, if you create a RPA bot  to prompt employees to change their passwords periodically for security reasons, RPA can consistently send notifications to workers based your company’s security policies.

Cognitive Intelligence

True Artificial Intelligence (AI) solutions require training and customization in order to provide businesses with maximum benefit. AI systems must integrate with all front- and back-end data, and also master business processes, industry terminology, and the measured reactions necessary to show empathy. These solutions are far superior to chatbots because they’re capable of learning, making their own decisions based on data (within pre-determined guidelines), and executing tasks.

A chatbot doesn’t understand context regardless of ambiguity or language variations. Chatbots can’t disambiguate similar intents. Chatbots can’t switch between multiple contexts because the context is what guides decision tree. Conversely, cognitive intelligence-based systems don’t follow such rigid decision trees, which means they’re able to jump back and forth during conversations in order to help customers according to how a customer guides the conversation.

Cognitive intelligence-based tools are capable of unsupervised learning. If a cognitive tool needs human help to complete a task, it will study the process the human used to complete the task, and master that process. It will then ask a human automation supervisor if it can apply this process to similar future requests.

Cognitive Intelligence and RPA

When you combine the best aspects of cognitive intelligence with RPA, even the least tech-savvy business user can create basic process automations. An employee can interact or converse with the RPA platform via cognitive intelligence, which guides the user through each step of the automation recording process. Once the process has been recorded, the intelligent RPA solution builds the automations for the user.

Intelligent RPA tools, such as 1RPA in IPsoft’s AIOps platform, are capable of understanding business processes and reading text through Natural Language Understanding, so they can process web pages, emails, programs, or any unstructured text you put in front of them-- just as a human would. Data is automatically organized according to business process requirements.

The Next Evolution of Automation

So what is the apex of the automation lifecycle? Put simply, it is an on-demand IT and shared services solution that communicates with employees to help them resolve basic IT and back office requests, such as the AIOps platform. Users can connect instantly with internal corporate systems, improving speed and productivity. Whether it’s resetting a password, logging vacation days, or filing an expense report, the system can provide services and support when needed.

Additionally, next-generation automation connects all of your IT operations tools and overlays them with cognitive intelligence so that no IT tasks get left out of the equation. The final result is end-to-end IT operations backed by intelligent automation. Your platform will automate even the most complex business processes, it will learn from user interactions and human inputs, and it will even identify and create new automations to improve processes. This ability to automate the creation of automations leads to more streamlined operations.

Determining where your business currently stands on the automation lifecycle, and how automation solutions can help you achieve your company’s goals and objectives, is an important consideration to ensure that such investments deliver value.


Previous Next

A Beginner’s Guide to Conversational AI

Crossing the bridge between digital assistants/chatbots and real Conversational AI requires a fuller understanding of how the technology works and its potential business value.

In our latest white paper, A Beginner's Guide to Conversational AI, we explore these subjects for companies pursuing a near- or long-term technology strategy that includes Conversational AI solutions and Digital Employees.

Download our white paper to learn how to generate business value with Conversational AI.