UiPath Announces New Platform Features

UiPath announced several new features in its platform which infuse GenAI deeply into the UiPath Business Automation Platform to help businesses achieve greater outcomes with AI and automation.

UiPath Autopilot for developers and testers

In the digital enterprise, rapid development and testing speeds are a top priority. UiPath delivers on the promise of AI-powered productivity with the release of UiPath Autopilot for developers and testers.

• UiPath Autopilot for developers uses the power of GenAI and natural language processing (NLP) in UiPath Studio to create workflows, generate expressions, and help build automations. With an over 70% acceptance rate, Autopilot for developers’ text to workflow is making things easier for less experienced developers and speeding up scaffolding for experienced developers. It delivers:
• Text to workflow: developers can simply describe an automation idea in natural language, and Autopilot will create the initial workflow.
• Text to expressions: with Autopilot, developers no longer need to remember the exact syntax and structure of expressions. They can describe what they need in natural language and allow AI to generate the correct expression.
• Text to code: generate code from natural language descriptions to shorten time to deployment for automation projects.
• PDF form, text, or image to UiPath Apps: in UiPath Apps, Autopilot enables developers to quickly create a new app—either from a PDF file or from an image of a paper form.
• UiPath Autopilot for testers accelerates every aspect of software testing by leveraging GenAI to refine and improve requirements, generate step-by-step tests from those requirements, and uses those tests to create coded automations. It provides:
• Quality checks: Autopilot analyses and evaluates requirements to ensure they are clear, consistent, and complete.
• Test design: from those requirements and supporting docs, Autopilot generates manual tests with step-by-step instructions.
• Test automation: Autopilot converts any text, such as manual tests, into automated tests in UiPath Studio with little to no effort.
• Test insights: Autopilot provides real-time, actionable insights into the test case portfolio to offer continuous improvement.

UiPath plugin and integration with Copilot for Microsoft 365

Now in preview, the integration enables joint customers to automate end-to-end business processes with co-workers directly within Microsoft Teams. Customers will have access to a prebuilt automation library to run automations that complete common, repetitive tasks, along with specialised automations for function- or industry-specific tasks. Users can also discover and run automations their company has developed.

New AI enhancements for IDP

UiPath is committed to building the best-in-class AI models for core business tasks. At UiPath On Tour London, the company released new AI enhancements for intelligent document processing (IDP) capabilities, including:

• UiPath DocPath and CommPath: specialised large language models (LLMs) enabling out-of-the-box processing of any document or message, with increased accuracy, speed, and scale. Both LLMs are now in general availability.
• Active learning: reduced training effort by 80% with the introduction of active learning techniques. Now, anyone can easily train their own models with active guidance from AI; no machine learning or coding skills are required.
• Generative validation: use GenAI to validate Document Understanding model outputs. Generative validation offers a “second opinion” on the data extracted from documents, reducing the workload of human annotators. This reduces the time spent on human validation, decreases average handling time, and boosts automation rates by up to 200%.

AI-powered building blocks

UiPath is using AI to simplify and improve higher-level automation tasks. GenAI activities are a curated collection of activities that simplify common GenAI use cases for automation developers, including:

• Content generation: generates a chat response for the provided request using chat completion models.
• PII filtering: scans the input text, identifies any Personally Identifiable Information (PII) or Protected Health Information (PHI), and replaces such data with a redaction marker.
• Summarise text: summarises text using an LLM.
• Translate: detects and reports the language of any given text.
• Object Detection: identifies and locates objects within images for tasks like automated inspections or data collection.
• Image Classification: categorises images into predefined classes to support processes like sorting products or analysing documents.
• Categorise: organises data into categories for improved analysis and decision making in workflows.
• Named Entity Recognition: extracts specific entities (like names, places, and dates) from text, aiding in information retrieval and organisation.

In addition, Intelligent UI form-handling capabilities leverage UiPath specialised AI and the learnings from Clipboard AI. These activities accelerate development and reduce maintenance as they deal with a form as a single entity, rather than a collection of individual fields. They include:

• Form fill: uses AI to complete a form. Given input data and a form, the activity will use AI to map data to the available input fields, fill in the data, and submit the form.
• Form copy: uses AI to read the data from a form. Given a target form, the activity will use AI to extract data and map it to a derived data structure.
• Set value: uses AI to input data into a single field.

All activities use the UiPath Trust Layer, ensuring customers can use GenAI-powered capabilities with the confidence that company data and policies govern them.