UiPath is firmly established as a leader in AI-powered automation. At this week’s TOGETHER conference in London, which UiPath bills as Europe’s largest automation and AI conference, the company unveiled new and updated features that significantly extend its AI functionality for customers.
To understand the shifting sands of how mid-market and enterprise CXOs are making purchase decisions to modernize technology, consider Acceleration Economy’s “Selling to the New Executive Buying Committee,” a Course designed to assist vendors, partners, and buyers in this process.
The announcements from UiPath provide both general and targeted benefits to users. We’ll run through each of the new and updated features, giving context to some lesser-known technologies integrated into the UiPath platform.
Generative AI Focus
UiPath is expanding the functionality of its core Business Automation Platform by introducing new generative AI capabilities. First, the company announced general availability of its OpenAI and Azure OpenAI connectors with support for GPT-4, the latest in OpenAI’s staggeringly successful GPT foundation. The connectors extend automation deeper into business operations and enable customers to develop actionable insights from their data.
Which companies are the most important vendors in AI and hyperautomation? Check out the Acceleration Economy AI/Hyperautomation Top 10 Shortlist.
In addition, UiPath will now support the Falcon Large Language Model (LLM) through the Amazon SageMaker connector and the preview of the Google Vertex connector, which includes support for PaLM 2, Google’s most advanced LLM. PaLM 2 is capable of handling a wide range of generative AI use cases including coding, classification, translation, and answering questions.
While many of the technologies adopted by UiPath in these latest updates may be familiar, Falcon may not be as familiar. Developed by the Technology Innovation Institute (TII), Falcon is a foundational LLM constituting 40 billion parameters.
While other LLMs include a larger range of parameters, Falcon boasts a series of other benefits, mainly in the realm of compute utilization. Comparatively, Falcon uses 75% of the training compute used by GPT-3 so it’s less resource intensive. TII also points out that Falcom LLM’s data pipeline includes a range of filtering and deduplication techniques when extracting information from the Internet for high data quality at scale.
Generative AI for Specialized Use Cases
UiPath offers over 70 specialized AI models for use cases including task mining, document processing, and operating unique data sets. Now, UiPath has added GPT-powered functionality to its existing Document Understanding and Communications Mining features to increase the efficiency of these specialized AI tools and make it faster to deploy tailored models.
For Communications Mining, GPT generates label recommendations to expedite the model training process. Classification and pre-labeling are improved in Document Understanding using the same technology.
“AI technologies are at the core of the UiPath Business Automation Platform, and we are excited to extend the capabilities of our enterprise automation with new Generative AI and Specialized AI offerings,” said Graham Sheldon, Chief Product Officer at UiPath. “UiPath is uniquely able to combine our understanding of screens, documents, tasks, and processes with the intelligence of Generalized AI to watch work happen, understand what is being done, and automate it in our platform. “
Toni Witt, Acceleration Economy practitioner analyst and AI/Hyperautomation Expert, noted two major benefits from UiPath’s new products and initiatives. “UiPath’s new Generative AI features dramatically reduce the time-to-value of using cutting edge models like GPT-4 for the company’s existing customers and prospective customers,” Witt says.
“It’s interesting to see that UiPath announced support for not just OpenAI models, but also Falcon, the base model of Amazon, and PaLM 2 from Google. This gives their customers the flexibility to choose between multiple base-level providers to meet their own needs in terms of cost, capability, and internal values around responsible AI development.”
UiPath also used the TOGETHER conference to announce the preview availability of Clipboard AI. Designed for finance teams, the new feature uses UiPath AI Computer Vision and generative AI technologies to simplify and speed up data transfer between documents, spreadsheets, and applications, understand the data in context, and automatically insert the data in the correct location. This new functionality will collectively help finance teams leverage UiPath automation to streamline the management of invoicing tasks better.
UiPath’s mission is to deliver efficiency across the enterprise. With these new and updated features, the company is underscoring this commitment while providing users with new functionality that delivers scalability and accuracy in tandem.
By choosing to invest in various LLMs, each with unique capabilities and strengths, UiPath is addressing diverse customer needs with the goal of optimizing its business processes across the board.