In brief:
- Artificial intelligence (AI) offers vast potential to transform the way we work, but as many as 80% of AI projects fail.
- Webinar attendants ranked identifying the right use cases, lack of clarity into the AI technology landscape, and skills gaps as the biggest roadblocks to successful implementation.
- Intelligent Document Processing tools are the most common entry point into AI that sets the stage for success.
- UiPath customers benefit from several unique capabilities that can turbocharge an AI journey, ranging from seamless integration of a wide range of AI tools to human-in-the-loop machine learning.
The emergence of Generative AI (GenAI) kicked AI adoption into high gear: the number of organizations infusing AI into at least one business function jumped from 55% in 2023 to 72% this year (McKinsey “The State of AI in Early 2024” report).
But while artificial intelligence offers vast potential to transform the way we work, most AI projects fail – with some estimates placing the rate of failures at 80%, a 2023 Harvard Business Review report states.
Strategies and use cases for closing this gap between enterprise AI goals and achievements were the key focus of Auxis’ recent webinar: “Take your UiPath Program to the Next Level with AI at Work.” Featuring UiPath Head of AI Strategy Ed Challis, ELS Properties Senior Manager of Business Applications Leah Hatcher, and Auxis Intelligent Automation leaders Eduardo Diquez and Bill Woodford, the webinar explored AI opportunities and challenges in the practical world, where most organizations struggle to scale automation and achieve expected ROI.
To view the full webinar, click here.
“Until now, it’s been very difficult for organizations to deploy AI models effectively,” said Auxis Intelligent Automation Director Eduardo Diquez. “It’s a very exciting moment to be dabbling in AI.”
Intelligent Document Processing: The most common AI entry point
What’s preventing enterprises from getting AI-powered automation right?
More than 70% of webinar attendants surveyed by Auxis listed the ability to identify the right uses cases as the biggest barrier to success, followed by lack of clarity into the AI technology landscape (46%) and skills gaps (42%).
“We see many clients for whom AI is a shiny object and they tend to lock in out of the gate on a particular solution,” said Auxis Senior Managing Director of Digital Transformation Bill Woodford.
Instead, Woodford advised companies to spend time evaluating the complexity, opportunities, and value the business can achieve from different solutions before jumping in – and then prioritize a roadmap accordingly. “This will 100% improve business outcomes,” he said.
To help organizations get started, webinar panelists detailed the most common AI journey that sets the stage for success. With technology that solves a universal pain point and is easy to understand, Intelligent Document Processing (IDP) solutions like UiPath’s Document Understanding and Communications Mining tools stand as great entry points into next-gen capabilities, advised Ed Challis, who spearheads AI strategy for the No. 1 business automation platform UiPath.
Document Understanding automates the extraction, interpretation, and processing of strategic information from high volumes of documents – from structured files like spreadsheets to unstructured documents like scans of receipts and handwritten forms. Communications Mining analyzes the content of various communication channels like emails or chats to uncover insights and automate tasks, and is a newer capability in UiPath’s portfolio.
“There’s clarity around how IDP technology works, what it does, and the use cases, so the business case can be very clearly established. That’s a great way to start to introduce AI to your business.”
– Ed Challis, UiPath Head of AI Strategy
Challis explained how rapid advancements in AI technologies like machine learning (ML) enabled IDP tools to move beyond the limitations of the original optical character recognition (OCR) technology – accurately processing complex documents, understanding context for data the automation needs to extract to improve results, and greatly shortening model training times for rapid deployment.
As demand for AI continues to surge, the global IDP market is projected to grow at a whopping 30.5% CAGR, expanding from $2.3 billion in 2024 to $19.32 billion by 2032.
Leading independent research firms like Gartner and Everest Group rank UiPath as the No.1 document processing solution – citing its ease of use, ability to process a wide range of documents, consistent accuracy, greater scalability, and speed to ROI. Its AI-powered Document Understanding tool has a proven ability to reduce extraction errors by up to 76% compared to other leading GenAI-powered models, according to webinar data.
“For anyone who has tried to use IDP in the past and they found that project to be slow, or maybe it didn’t yield the kind of ROI they wanted, GenAI makes IDP projects much, much faster, and much, much easier to implement,” Challis said. “No matter where data arises, if it has a role in a business process, our goal is to be able to process it.”
Leveraging UiPath to accelerate your AI journey
Many use cases require different AI skills, and UiPath’s unique ability to seamlessly integrate and orchestrate a wide range of AI systems is what takes its offerings to the next level, Diquez said.
Combining IDP with RPA and Generative AI, for example, allows organizations to implement AI-powered solutions that can action complex, end-to-end workflows like claims processing, financial reporting, and customer onboarding. GenAI acts as the “brain” – classifying documents and helping IDP tools mine unstructured forms. Robotic process automation is the “muscle” – gathering relevant current and historical data to create the context IDP needs to produce the most accurate results and uploading extracted information.
“Context is the secret sauce that allows GenAI models to work effectively – without it, the results are mediocre. RPA is a key component to driving AI value.”
– Bill Woodford, Sr. Managing Director of Digital Transformation
But orchestrating the range of AI interactions and integrations required to effectively solve most use cases is one of the greatest challenges when transitioning from pilot to scale (McKinsey 2024 “Moving Past GenAI’s Honeymoon Phase” report). Without leveraging a comprehensive platform like UiPath, enterprises are forced to implement, maintain, and manage separate environments and licenses for different AI models – significantly increasing complexity and failure rates, Diquez explained.
UiPath provides access via connectors to all the leading GenAI models, including OpenAI, Azure Microsoft OpenAI, Anthropic, and more. That offers organizations the much-needed ability to test different GenAI solutions to find what works best for each use case, instead of being locked into a single model.
“Usually with a use case, you’re not just using one simple model to extract or get to the answer, you are using a combination,” Diquez said. “Being able to quickly deploy different models securely and effectively, with very limited overlay on infrastructure and integration needs, is very exciting for UiPath customers.”
UiPath’s customers further benefit from the ability to integrate solutions like AI-powered document processing into the platform’s broader automation workflows. Strategic human-in-the-loop (HITL) capabilities provide governance when human review or decision-making is required – allowing bots to identify tasks that require human intervention, assign them to staff, and seamlessly resume the rest of the automated process once the exception is resolved.
HITL also drives performance improvement, enabling AI models to continuously learn as people make decisions and act on process exceptions.
“If you’re trying to automate responses from human resources and someone asks a finance question, GenAI will create an answer without the right guardrails. Being able to raise your hand and pass those questions that you don’t want GenAI to answer to a human makes for a very complete, strong, enterprise-ready AI deployment,” Diquez said.
AI priorities vs. success rates: What’s the best path forward?
IDP and Process/Task Mining rank as the AI technologies webinar attendants are prioritizing most in 2024, according to Auxis poll results. Process/Task Mining provides real-time insights aimed at helping organizations improve their business operations.
Yet only 10% of poll respondents said they had currently implemented either of these technologies successfully. Organizations are also interested in GenAI, with 36% of webinar attendants prioritizing implementation this year. But only 5% said they have deployed GenAI effectively.
Unsurprisingly, RPA showed the highest rate of successful implementation. Widely adopted over the past decade across business processes, RPA has been successfully implemented by 71% of Auxis poll respondents – and 44% expect to continue prioritizing it this year.
Diquez said lack of AI skill sets is often to blame for the discrepancy. CFOs cite talent as the biggest hurdle to GenAI adoption in finance departments, with GenAI technical skills (65%) and fluency (53%) the most pressing concerns.
The newness of AI technology combines with the tight labor market to make it challenging and expensive to hire and retain highly skilled talent in-house. Not surprisingly, organizations are increasingly turning to intelligent automation providers like Auxis with the resources, knowledge, and expertise to identify the right opportunities, prioritize a roadmap, and achieve sustainable AI success.
“The skill sets that drove RPA development are not the same as being able to build IDP models. We constantly receive requests from organizations that have mastered RPA and they’re looking to add AI skills to their Centers of Excellence.”
– Eduardo Diquez, Managing Director of Intelligent Automation
A UiPath Platinum Partner, Auxis maintains 150 developers at its Automation Center of Excellence in Latin America who are certified across the breadth of solutions that comprise UiPath’s comprehensive business automation platform. Leveraging this deeply experienced talent pool has helped companies across industries power successful AI and automation journeys – combining significant labor arbitrage and highly skilled resources with nearshore advantages like real-time collaboration, outstanding English fluency, and cultural alignment that allow Auxis to function as a seamless extension of your team.
An AI use case: Creating human-like, personalized user interactions
During the webinar, ELS Properties’ Leah Hatcher detailed how her company partnered with Auxis to innovate an AI solution that transformed its ability to quickly respond to internal and external user interactions.
A leading property management company in North America, ELS (aka Equity LifeStyle Properties) receives thousands of internal and external user interactions online – and replying manually was a time-consuming and laborious process. But while the company was interested in using AI to solve its challenges, it was not in favor of using tools which would generate generic, robotic messages.
By deploying a combination of UiPath RPA and GenAI, Auxis helped the client achieve 95% touchless, real-time, human-like, personalized messages in multiple languages. The result: greater user satisfaction, better insights into market preferences, and a stronger position for ELS in the market.
Hear ELS Properties Senior Manager of Business Applications Leah Hatcher describe how partnering with Auxis enabled ELS to leverage AI and automate user interactions effectively.
Challis said ELS’ experience aligns with the first of three main patterns for GenAI-powered automation use cases that have emerged to date:
- Reader/writer – Automation gathers context from multiple sources to help GenAI generate personalized messages that the automation distributes. Use cases include user feedback response, compliance reporting, drafting customized letters, product documentation, and more.
- Analyst/doer – Automation validates structured outputs from large language models (LLMs) and executes processes based on the content. Use cases include multi-source report creation, reconciliations, data extraction, vendor selection, competition analysis, and more.
- Assistant – Automation adds context and action to conversational assistants. Use cases include knowledge base assistant, self-service help desk, employee travel concierge, supply chain buyer assistant, and more.
Embracing AI at Work: Why status quo is no longer an option
While many businesses have taken a cautious approach to emerging technologies like GenAI, webinar panelists agreed that waiting and watching is no longer an option for maintaining a competitive edge.
Nearly 70% of organizations that moved beyond pilot stages expect to see meaningful results from their GenAI initiatives in the next 12 months, according to Dell’s 2023 Generative AI Pulse Survey.
“While automation enabled the transformation of business processes, AI takes it to the next level – enabling human-like activities and turbocharging productivity in ways that we have never been able to achieve before,” Woodford said. “The pace of technology change is accelerating and there is major risk in staying with the status quo.”
View the full “Take Your UiPath Program to the Next Level with AI at Work” webinar to unlock more strategies for transforming your business with AI from our AI and automation experts.
Want to learn more about how Auxis can help you leverage UiPath IDP, AI, and Generative AI for business transformation? Schedule a consultation with our intelligent automation experts today! Or, visit our resource center for intelligent automation tips, strategies, and success stories.
