Automation Anywhere levels up business process automation with generative AI

Business process automation software firm Automation Anywhere Inc. used its annual user conference today to unveil a new AI + Automation Enterprise System that infuses the latest advances in generative artificial intelligence with its products.

The new offering was announced at Imagine 2024, and is built around a series of new AI agents built using the company’s own AI Agent Studio platform. The company says these agents accelerate process discovery and speed up the development and deployment of more powerful automations that can increase business efficiency and drive significant productivity gains and cost savings.

Automation Anywhere is a leader in a field known as robotic process automation or RPA, which is AI-powered software that’s designed to mimic the actions of human workers. RPA tools are used to automate mundane and repetitive business processes such as data entry, so humans can focus on higher-value tasks.

The company says the new AI + Automation Enterprise System is designed to take automation to the next level. The new enterprise AI agents can create more advanced RPA bots that handle complex tasks requiring greater cognitive skills. They learn from enterprise data, make appropriate decisions and take action across any enterprise system of record, the company explained.

Automation Anywhere co-founder and Chief Executive Mithir Shukla (pictured) said every enterprise today is feeling under pressure to work smarter not harder, and be more productive and efficient. “Automation has gotten us part of the way there,” he said. “AI Agent-powered automation is the breakthrough that will take us beyond that, to automate the seemingly impossible.”

Next-gen AI Agents to accelerate automation

What’s different about Automation Anywhere’s next-generation enterprise AI agents is that they’re infused with generative AI, which gives them the ability to expand process automation beyond rules-based work. They combine AI and action to take on more complex tasks, such as identifying and replacing products in case of a stock shortage. What’s more, they can learn and adapt as they mature, so over time they will be able to automate such complex tasks more rapidly.

Customers can design their own automations using the new AI Agent Studio, which is a low-code environment for building, managing and governing customized generative AI agents that can handle cognitive tasks. Users simply choose the foundational model they require, connect it to their dataset to enhance its knowledge with context, and package it with pre-tuned prompts to ensure its outputs are relevant for the task in hand.

The company has also built what it says are enhanced security and governance controls to protect corporate data and monitor the outputs of those models. Those tools include guardrails to ensure consistency, and prompt testing to limit AI hallucinations, which is when AI fabricates its responses.

Besides the AI Agent Studio, customers can also leverage pre-built AI Agents such as the new Document Automation agent, which can process any type of document in real time, including unstructured documents. It can extract and summarize information, including data from complex tables embedded within those files, then use the insights it gathers to perform various tasks, the company said.

Enhanced automation workflows

Also announced today was a new Automator AI tool that’s said to speed up automation development lifecycles. The underlying engine of this tool is powered by the company’s proprietary generative AI process models, and can be enhanced with third-party large language models and fine-tuned with additional data.

Automator AI’s capabilities include an autopilot that can transform natural language commands into automation process flows, cutting development times by as much as 30%, the company said. It enables users to go from discovery to automation in double-quick time, rapidly converting process documentation into a draft automation powered by generative AI.

In addition, there’s a new Automation Co-Pilot, separate from Automator AI’s autopilot, designed to handle conversational automation requests from users. Available in preview, it has been integrated with Amazon Q to provide real-time, on-demand assistance to help workers find the information they need or request automations for specific tasks. Because it’s conversational, it will understand users’ requests and provide the answers they’re looking for, or suggest an automation they can run, for any application it’s embedded with.

Last, the company announced a new Service Operations Solution Accelerator that’s available now to help teams get started with pre-packaged AI agents for a range of service operations, such as order management, returns processing and customer service questions and answers.

The updates announced today will help Automation Anywhere’s customers get more out of their existing AI investments, International Data Corp. analyst Maureen Fleming believes.

“These latest enhancements… make AI-powered automation even more accessible, empowering organizations and employees to leverage AI in new ways, helping them to work smarter than ever before,” she said.

Shukla appeared on SiliconANGLE Media’s mobile livestreaming studio theCUBE in April, during its coverage of Google Cloud Next 2024, where he discussed the numerous ways in which generative AI is transforming automation:

Photo: SiliconANGLE