Integrate GenAI capabilities within UiPath: Demos and Use Cases
- 1. 1 Integrate GenAI capabilities within UiPath: Demos and Use Cases
- 2. 2 Safe Harbor This presentation may include forward-looking statements. Forward looking statements include all statements that are not historical facts, and in some cases, can be identified by terms such as “anticipate,” “believe,” “estimate,” “expect,” “intend,” “may,” “might,” “plan,” “project,” “will,” “would,” “should,” “could,” “can,” “predict,” “potential,” “continue,” or the negative of these terms, and similar expressions that concern our expectations, strategy, plans or intentions. By their nature, these statements are subject to numerous risks and uncertainties, including factors beyond our control, that could cause actual results, performance or achievement to differ materially and adversely from those anticipated or implied in the statements. Although our management believes that the expectations reflected in our statements are reasonable, we cannot guarantee that the future results, levels of activity, performance or events and circumstances described in the forward-looking statements will be achieved or occur. Recipients are cautioned not to place undue reliance on these forward-looking statements, which speak only as of the date such statements are made and should not be construed as statements of fact. This meeting is strictly confidential. By participating in this meeting, you agree to keep any information we provide confidential and not to disclose any of the information to any other parties without our prior express written permission. Neither the information contained in this presentation, nor any further information made available by us or any of our affiliates or employees, directors, representatives, officers, agents or advisers in connection with this presentation will form the basis of or be construed as a contract or any other legal obligation.
- 3. 3 Senior Automation Developer Boundaryless Group Stefano Negro Hosts Technical Account Manager UiPath Flavio Martinelli
- 4. 4 • Introduction • UiPath GenAI Activities • GenAI Activities Package • Live Demos • Q/A Agenda
- 6. 6 Our Journey to AI-powered Automation RPA UI Automation Computer Vision API Automation Low-code Apps Orchestration Governance Cloud Discovery Analytics Continuous Testing Intelligent Document & Communication Processing 2023 2018 Generative AI Powered Automation “ UiPath AutopilotTM Robotic Process Automation application Enterprise automation suite End-to-end platform for AI at work 2020 UiPath innovates at an incredible pace, prioritizing our most important needs for success, from security to scalability. Chief Automation Officer Global Media and Digital Communications Leader UiPath LLMs UiPath GenAI Activities
- 7. 7 AI is surpassing human performance 2000 2002 2004 2006 2008 2010 2012 2014 2020 100% (Human performance) 80% 60% 40% 20% Source: 1998 HANDWRITING RECOGNITION IMAGE RECOGNITION SPEECH RECOGNITION 2016 READING COMPREHENSION 2018 LANGUAGE UNDERSTANDING COMMON SENSE COMPLETION CONTENT GENERATION 2022 CODE GENERATION
- 8. 8 AI-powered Automation Products Generative AI Specialized AI Open | Flexible | Responsible Action AI Trust Layer UI API Human in the Loop Context Documents Data Processes People Communications
- 9. 9 Our approach to Generative AI Supercharge Productivity 1 Enhance Quality & Accuracy 2 Automate More 3 GenAI unlocks new Automation use cases. Upskill talent and supercharge productivity. Reduce friction with increased accuracy, performance and reliability.
- 10. 10 Enhance Quality & Accuracy Generative Extraction & Validation Zero-Shot Experiences Specialized Models 70+ pre-built models to analyze and process documents and communications across industries and domains. Active Learning Active learning employs an iterative process between annotators and model greatly reducing the time and effort required for both annotating and training. Zero-shot training automatically suggests label names for message groups, creates those labels and trains a first model. Identify multiple requests, extract the relevant data and, crucially, understand the relationships between them. Reduce friction with increased accuracy, performance and reliability. Enhance Quality & Accuracy
- 11. 11 What is Context Grounding? Joe has questions around GenAI at UiPath What are the topics in the GenAI Security Handbook? • Data Security and Compliance • Risk Mitigation • Human-in-the-Loop Systems • Specialized LLMs LLM Response 3rd Party LLM Joe has questions around GenAI at UiPath What are the topics in the GenAI Security Handbook? • UiPath AI Trust Layer • End-User and Server Security • Integration Service • Comm Mining and Document Understanding GenAI Security Handbook 3rd Party LLM LLM Response Context Grounding: A process that ensures generated responses are relevant and tailored to your specific situation
- 12. 12 Context for UiPath GenAI Query Context Response Embed queries and get relevant semantic context Parse and reason contextual information to construct responses Iterative feedback loops in Action Center, Assistant and Autopilots Retrieval Augmented Generation: • Overcomes (existing) LLM context window limitations and optimizes interactions with large context windows • Reduces risks of hallucination through reference to ground truth data stores • Gives Generative apps access to specialized and proprietary knowledge sources • Gives Generative apps access to up to date sources of information • Enables positive feedback loops between data stores and user-queries
- 13. 13 Unreliable Predictions Stagnant Insights Timely, Costly Adjustments Without Context Grounding Bring business context to your automations Accurate, Consistent Predictions Dynamic, Flexible Knowledgebases Relevant AI On-Demand With Context Grounding Business Data Context Grounding Automation Developer UiPath GenAI Experiences LLM-Ready Business Context UiPath-Managed LLMs No data retention or training Human Validation UiPath Trust Layer
- 15. 15 Major Benefits Ready-to-use activities for common GenAI tasks that leverage industry-leading vendor models Pre-tested, quality controlled prompts, that are monitored according to internal relevancy benchmarks Pre-tested, quality controlled prompts, that are monitored according to internal relevancy benchmarks Empowers users to build custom dynamic prompts that can pull any application data as needed Injects proprietary context without requring the management of complex architecture Works without requiring any third-party LLM subscriptions
- 17. 17 • https://community.uipath.com/Torino • https://community.uipath.com/Rome • stefano.n@boundaryless.com • flavio.martinelli@uipath.com Questions?