Adobe Systems Inc. today launched two new cloud services for building robotic process automation workflows and applications that require the ability to process documents.

The new services give Adobe another way to address the rapid adoption of RPA and machine learning among its enterprise customers. 

Enterprises are investing in automating repetitive business tasks as part of their digital transformation initiatives. Many of the business tasks being automated are chores that involve processing documents, such as preparing contracts and copying data from invoices to a company’s accounting database. Adobe says its new services can boost such projects. 

Both services, called Adobe PDF Extract API and Adobe Document Generation API, respectively, are focused on automating tasks that involve documents created in the PDF format. A sizable portion of the world’s business documents are stored in the format. According to Adobe, an estimated 2.5 trillion PDF files are created every year.

PDF Extract API, the first service introduced today, uses machine learning to extract documents’ contents. It’s not the first offering on the market with that capability. But Adobe says that whereas existing offerings often focus on extracting only one specific type of content from documents, PDF Extract API can retrieve all the elements in a file, including not only text but also other items such as embedded spreadsheet tables.

The company sees enterprises using PDF Extract API in robotic process automation projects. RPA is the term for a category of software products that use artificial intelligence to observe how employees perform a task in a business application and then create a software “bot” that can replicate the task. According to Adobe, PDF Extract API enables RPA bots to extract data from documents and incorporate the data into the tasks they perform. 

Copying information from PDF invoices to a company’s accounting database is one example of how the technologies may be used together. According to Adobe, RPA bots can perform more fine-grained tasks as well with the help of its new service. A bank could create an RPA bot that extracts the content of forms submitted by customers and, based on the information in each form, automatically determines the department to which it should be sent. 

Another use case that Adobe is targeting is artificial intelligence training. To build an AI application that performs a certain task, a company’s developers must find a suitable neural network and then supply it with training data from which it can learn how to perform the task. There are scenarios where that training data is not readily available, but must be extracted from a company’s documents.

A software maker building an AI chatbot to answer product questions, for example, may wish to train the chatbot on product documentation from its knowledge base. If the product documentation is stored in the PDF format, the software maker could use PDF Extract API to fetch the relevant information for its AI. 

Adobe says that the service can not only extract the text and other elements in a document but also glean contextual details, such as if a certain piece of text is from a paragraph or a footnote. This context can simplify AI development in certain respects. It allows companies to avoid extracting content snippets such as footnotes that may not be useful for AI training. 

“Organizations can use PDF Extract API to quickly and accurately extract data for use in machine learning models, analysis, indexing or storage,” said Vibhor Kapoor, the senior director of marketing and go to market for Adobe’s Document Cloud business. 

Companies can also use the service to “republish PDF content across different media,” the executive explained. The context that PDF Extract API gleans about each content element, such as if a piece of text is part of a footnote, makes it easier to replicate the original layout of a document when it’s republished in a new medium. 

The other service Adobe introduced today is Document Generation API. It allows enterprise developers to create applications that automatically customize documents and fill forms.

In the business-to-business segment, companies usually onboard new customers using ready-made sales contracts that they customize for each client. This customization can be time-consuming in complex deals because a contract may contain a lot of information that is unique to a specific customer. Document Generation API automatically retrieves the needed information from a company’s customer relationship management platform and other business applications, then add it to the contract.

The service can be used together with PDF Extract API, the first service Adobe debuted today. A company could automatically create contracts and other documents using Document Generation API, then make the file available to RPA bots for further processing via PDF Extract API. 

To simplify such projects, Adobe is also rolling out an integration with Microsoft Corp.’s Power Automate. That’s an RPA product that can perform repetitive business tasks to free up time for a company’s employees.

“The Document Generation API is also available with the Adobe PDF Services connector in Microsoft Power Automate, enabling Power Platform users to easily automate the preparation of their documents for invoices, agreements and more,” Kapoor wrote. “These users can also access 15 new, ready-to-use templates for document workflows in Power Automate.”

The new services expand Adobe’s Document Services suite, a bundle of PDF-centric automation services for developers that it introduced last year. The market reception appears to be positive: Adobe says developer signups are up more than 80% quarter-over-quarter.

The Document Services suite gives Adobe a way to monetize the rapid growth of RPA and AI in the enterprise. The suite indirectly also helps boost the company’s lineup of tools for authoring PDF documents. The more ways there are for companies to extract value from their PDF files, the more likely they are to continue using the document format, which in turn means a bigger addressable market for Adobe’s PDF authoring tools.

In conjunction with today’s product updates, Adobe said it’s making the Document Services suite of which the two new services are part available via the AWS Marketplace to simplify purchasing for Amazon Web Services Inc. customers.

Image: Adobe