Robotic process automation, software that emulates how people carry out tasks in a process, is becoming one of the principal enterprise use cases for artificial intelligence.

Established RPA solution providers are becoming prominent players in the cloud application development arena. Chief companies in this segment — which include Automation Anywhere Inc., BlackLine Inc., Blue Prism Group, Kofax Inc., Pegasystems Inc. and UiPath — are blurring the already fuzzy lines among RPA, business process orchestration, Web content management, edge computing and application development. And most of them now emphasize the depth of their solution portfolios’ AI capabilities.

When we speak of AI in an RPA context, the discussion has traditionally been focused on the technology’s use in inferring an application’s underlying logic from artifacts that are externally accessible at the client level. In this regard, machine learning and other AI tools typically drive the screen scraping of user interface presentation elements, optical character recognition of onscreen text, autosensing of browser-level control and domain object models, capture of human-user keystrokes and clicks, understanding of natural-language text and parsing of document object models.

Within RPA ecosystems, AI’s role is beginning to broaden. Increasingly, AI also provides the nucleus of intelligent bot-embedded application logic that addresses particular business process automation scenarios. A key industry milestone in this trend was Automation Anywhere’s announcement Wednesday that it had opened its online Bot Store to partner applications that embed AA’s newly available IQ Bot AI capabilities. At its annual customer event, the company announced that its partners can now offer pretrained application-specific IQ Bots that intelligently automate such document-centric business processes as handling financial statements, purchase orders and invoices.

AA’s Bot Store provides a one-stop shop for a wide range of application-focused RPA bots, not limited to just those that incorporate these new AI capabilities. Until the release of IQ Bot technology, partners and developers could only offer AI capabilities through API-level integration with external AI applications via AA’s Enterprise RPA platform. The introduction of IQ Bot enables developers and partners to offer native computer vision, natural language processing and ML capabilities in their bots. Already, several partners and developers — including Google Inc., Microsoft Corp., IBM Corp. and Pure Software Inc. — have contributed pretrained “cognitive” bots to the store.

Essentially, these AI-powered bots enable RPA applications to embed probabilistic business logic in conjunction with or in lieu of deterministic business rules. This innovation opens AA’s application ecosystem to the intelligent, contextual and situationally aware runtime behavior afforded by ML, deep learning and other data-driven AI algorithms. Consequently, it’s now possible for bots in AA’s RPA environment to leverage big-data and real-time streams in order to dynamically adjust their automated actions to a business process’ full historical, current and predictive context.

Market trends

AA’s latest announcement is a further proof point for several application market trends that Wikibon called out in our recent Big Data Analytics Trends and Forecast:

• Application developers have adopted an AI-first go-to-market approach for differentiating their offerings.
• App vendors are shifting their solution portfolios toward delivery of packaged applications that embed pre-trained AI to deliver fast industry or task-specific business outcomes.
• Application platforms are being architected for frictionless feature evolution through continual refresh of the AI, metadata, rules, graphs and other intelligent artifacts that have been deployed to the cloud edge.
• Development tools are evolving toward a focus on building and orchestrating AI microservices that have been embedded in distributed application environments, especially in mobility, robotics, sensor networks and other edge scenarios.
• More new enterprise application-development projects that come online involve building AI-driven smarts for deployment to mobile, embedded and edge endpoints, as well as to domain-specific application platforms.

One might regard these new solution-focused IQ Bots as equivalent to containerized AI microservices of the sort that Wikibon covered in this report last year. Already, it’s possible for IQ Bots to be combined to automate more complex business process orchestrations with the real-time agility. Updating the bots involves downloading their latest pretrained versions from the Bot Store and redeploying them into the applications into which they’ve been configured.

One of RPA’s big advantages in legacy environments is that, by building applications from external interfaces, it requires little or no changes to existing information technology systems. Consequently, it’s an easy drop-in to application development and process design shops, allowing developers to boost their productivity by implementing lightweight orchestrations among built-up applications. Wikibon expects a fair amount of development to migrate to the edges of the business, in the form of full-fledged intelligent RPA-based AI microservices at the heart of every application initiative.

Going forward, we hope that AA will evolve its RPA ecosystem to support cross-bot orchestration using cloud-native approaches such as Kubernetes. Picture a world in which developers can simply transform an existing enterprise workflow into an orchestration of interactively intelligent RPA bots. Now imagine that the developer can work this same magic across distributed bot-based microservices in private clouds, complex multiclouds and even among edge nodes on the “internet of things.”

Wikibon expects that other RPA vendors will compete with AA’s new offering by introducing their own prepackaged, pretrained solution-oriented AI bots within the coming six to 12 months. By taking its pioneering move into this new segment, AA’s latest announcement will set the stage for a “razor-and-blades” approach to building a new RPA business model focused on prepackaged vertical and line-of-business solutions.

Here’s a good discussion by Automation Anywhere Chief Executive Mihir Shukla on the digital transformation potential of RPA technology:

Image: geralt/Pixabay