17 May 2018

A look at RPA and AI in action: two success stories

17 May 2018

A look at RPA and AI in action: two success stories

Robotic Process Automation (RPA) and Artificial Intelligence (AI) are able and useful technologies on their own: while RPA can streamline repetitive, rules-based business processes, AI algorithms can manage business tasks that require more complex decision-making and analysis. In fact, these two technologies are even more powerful when implemented together. The convergence of RPA with such “intelligent” technologies is on the horizon and paints an optimistic picture for the future of automation in the workplace.


In leveraging this partnership, businesses will be able to rely on a single software solution to aid complex decisions and increase their business agility, a crucial component for success in today’s competitive markets. Let’s take a look at two real stories to see how a global consultancy firm and an insurance company have leveraged UiPath’s Enterprise RPA platform in combination with AI technology to drive new efficiencies.


RPA and NLP in global finance audit delivery


A global audit, tax, and advisory consultancy firm’s delivery center performs the quarterly and annual audits for their end customers. The audit process is a complex one, covering multiple sub-processes including running footing and cross-footing, checking internal consistency, recalculating balance sheets, and performing previous year closeouts. The work required was initially manually driven and very time-consuming. What’s more is that the processing logic was purely judgment based, with pre-defined rules applying across the input financial reports.


UiPath’s Enterprise RPA Platform now automates many of the activities required for audit delivery, including monitoring 10 Ks and 10 Qs, footing and cross-footing, checking internal consistency, recalculating the balance sheet and previous year tie-outs. Using Stanford's natural language processing (NLP) capability, UiPath robots contribute to the automation of financial reports by standardizing the business process logic and validating data in tables and in the narrative. The robots process textual contents, extract tables utilizing a PDF extraction engine, and perform validation of the original balance sheet and other financial data. 


85% accuracy improvement

54,000 hours of work saved


Through this combined RPA-AI solution, the global consultancy firm was able to save 54,000 hours of work annually, achieve 45% automation of the audit process, and reach 85% accuracy improvement in comparison to the 25% improvements the company had previously managed to attain using IBM Watson.


Automating complex, non-routine insurance activities


South Africa’s largest privately owned insurance group, Hollard faced a volume of 1.5 million emails incoming on a yearly basis from its brokers. To handle each insurance claim, the content of these emails and their attachments had to be interpreted, classified, and processed. Difficulties were primarily associated with a growing backlog of unprocessed claims as well as with maintaining compliance with Service Level Agreements (SLAs) and specific regulatory and statutory provisions.


Hollard turned to intelligent solutions provider LarcAI to streamline this complex, non-routine activity, who chose UiPath for its uniquely open architecture that allowed for the integration of AI services needed for end-to-end automation. The implemented solution involved machine learning (ML), natural language processing (NLP), intelligent optical character recognition (OCR) and analytics capabilities from Microsoft, IBM Watson, and ABBYY, all blended into UiPath’s Enterprise RPA Platform.


UiPath robots now access and interpret the inbound email’s content contextually, categorize and sort the needed documentation, and enter the necessary data into relevant databases. The robots also interact with human employees to execute particular instructions and deliver completion confirmations.


600% faster processing91% lower cost per transaction


With new claims now processed in real time, automation eliminated Hollard’s backlog of cases and reduced transaction costs by 91%. The robots’ more-than-human accuracy also ensures that all compliance requirements are met. UiPath robots can deal with 98% of cases autonomously and process them 600% faster. As a result of these changes, Hollard has also been able to achieve drastically improved employee-broker relations. Read the  entire story here


Are we there yet?


Powerful RPA and AI use cases are beginning to surface, although adoption is in its very early phases. AI is costly, risky and for many executives still an alien topic. McKinsey & Company suggests that only 20% of enterprises use AI technologies at scale or within core parts of their businesses. Could RPA boost these low adoption rates?


UiPath’s approach to AI


UiPath’s approach to AI is twofold. We build as well as partner and integrate AI functionality into our product.


At the heart of UiPath is our AI enabled computer vision, which we’ve spent more than a decade perfecting. Computer vision equips our robots with an exceptional ability to detect screen elements with human-like precision, making UiPath much faster to implement than its competing solutions, a significant advantage that gets multiplied in Citrix environments.


We are also working on incorporating machine learning and natural language processing into our product to augment the capabilities of UiPath robots to deal with relatively uncomplicated business exceptions and other unstructured data.


The other way that we plan to make RPA the main gateway to bring in more AI into our customers’ organizations is through partnership. Since there is no such a thing as "universal AI', there are going to be countless numbers of specific AI applications developed by specialized companies. Our objective is to offer ease of integration between these and our platform and to create a space of exchange for where solutions and best practices can be shared


For all of the above, we have the most substantial research and development commitment across both attended and unattended robot operations in the world.


Would you like to know more about UiPath's intelligent OCR and NLP capabilities? Watch our webinar with ABBYY


Watch Webinar 


by Mina Deckard

Show sidebar