Intelligent Automation in the Federal Government: What, When, and How?

We increasingly get requests from federal agencies to help them add intelligence—artificial intelligence (AI), that is—to their Robotic Process Automation (RPA) software bots. They are excited about their core RPA installations and happy with all the productivity, efficiency, and cost benefits they’ve accrued. But although they can do a lot with traditional RPA using its built-in rules and logic, they want to do more. They want to put intelligent automation to work.

Many of them have begun doing just that. A recent survey found that 63% of federal agency executives thought they could benefit from intelligent automation and that 38% have already started implementing it. But a major insight of the survey is that three in 10 (30%) of federal agency executives didn’t know if intelligent automation could help them. To us, that means that some education is in order.

In this blog, we’ll cover the what, when, and how of intelligent automation. Based upon the webinar “Intelligent Automation for the Federal Government,” sponsored by Automation Anywhere and DataRobot, you can read about the intelligent automation efforts of three federal agencies to understand where they are and where they are going in their automation journeys.

What: defining intelligent automation

So what is intelligent automation?

Intelligent automation is what happens when you combine RPA and cognitive computing technologies such as AI, machine learning (ML), and optical character recognition (OCR). You free your bots from the limits of structured and predefined data and tasks. Intelligent bots can learn, and they can make decisions. In effect, we’ve moved from programming bots to teaching them. And in some cases, intelligent automation allows us to do things that we never thought possible because bots can analyze data so large that humans would not be capable of doing it themselves.

What are the benefits of intelligent automation? According to the Federal News Network survey, 73% of government agency executives said it would move employees away from redundant or low-value work. Following that, 60% said it would reduce errors and improve the consistency of their missions’ outcomes. And 59% said it would supply a better experience to their “customers” (citizens and other agencies). Note that reducing costs came down at the bottom, with just 44%. This shows that perceptions of automation have changed in the marketplace. It’s no longer seen as a cost-reducing job killer but a job enhancer and a strategic advantage for agencies.

Erica Thomas, RPA program manager for the U.S. Department of Defense, says, “As technology continues to grow, and we move into more cognitive areas like AI and machine learning, they’re going to broaden and open up opportunities for automation within our department.”

When: timing for intelligent automation

According to agencies that are in the process of implementing intelligent automation, it’s good to start first with basic RPA, and then, once you have that mastered, start adding intelligence and cognitive capabilities to it.

Christine Gex, leader of intelligent automation at the National Security Council (NSC), says her agency is focused on the customer experience. It’s integrating other technologies with RPA, such as AI-based digital agents and the Microsoft Teams platform, for its initial intelligent automation initiatives. “Right now, we’re mostly using it to reset passwords, but we’re also proceeding to other lines of business.

Rajeev Dolas, chief technology officer (CTO) for the National Highway Traffic Safety Administration of the U.S. Department of Transportation (DOT), says that his agency is at the very early stages of implementing RPA. Only after that is solid will the agency get into intelligent automation with AI and machine learning technologies. “We’re a very, very lean organization, and the volume of tasks on everyone’s plates are just so numerous that intelligent automation is going to play a key role as we move forward,” he says.

How: Automation Anywhere and DataRobot

Automation Anywhere, a leader in cloud-native, AI-powered RPA, has partnered with DataRobot, a pioneer in automated machine learning, to help federal agencies build successful intelligence automations. By integrating with DataRobot’s platform, Automation Anywhere’s native cognitive automation capabilities are extended to include intelligent decision-making without human intervention.

Agencies start with the Automation Anywhere Digital Workforce platform that already allows any business user to build bots that automate rote, repetitive work. Plus, the combination of RPA, cognitive, and analytics technologies embedded in the Automation 360 platform enables automating complex business processes—even those that use unstructured data.

In addition to that, machine learning models from DataRobot can be trained using this data and subsequently harnessed to solve real-world business problems.

Automation Anywhere and DataRobot work seamlessly together. First, Automation Anywhere sorts the data into the right format for machine learning. It is then loaded into the DataRobot system, which automatically trains a model. This model can then be fed back into an RPA bot to perform tasks.

Federal agencies today collect vast amounts of data for all sorts of government processes. By deploying Automation Anywhere and DataRobot together, agencies can further democratize and accelerate their digital transformation journeys.