UiPath Inc v. Rule 14 LLC

1. Case Identification

• Case #: IPR2025-00623
• Patent #: 10,108,679
• Filed: March 6, 2025
• Petitioner(s): UiPath, Inc.
• Patent Owner(s): Rule 14 LLC.
• Challenged Claims: 1-20

2. Patent Overview

• Title: Data Mining System
• Brief Description: The ’679 patent describes a data mining system that uses a query to search data sources like social media. The system selects a data source for monitoring based on a correlation between the query and search results, and dynamically adjusts parallel processors for data analysis based on the rate of data extraction.

3. Grounds for Unpatentability

Ground 1: Obviousness over Safford, Chen, and Laribi - Claims 1-2, 8-12, 15-16, and 19-20 are obvious over Safford, Chen, and Laribi.

• Prior Art Relied Upon: Safford (Patent 9,288,123), Chen (Patent 8,010,524), and Laribi (Patent 9,769,085).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Safford taught a system for analyzing social media signals associated with a brand (“topic of interest”), identifying relevant accounts, and using “collectors” to periodically poll and extract data from those accounts. Chen was cited for the obvious modification of including a brand name in Safford’s keyword query. The combination of Safford and Laribi was argued to render obvious the limitations concerning dynamic resource scaling. Safford disclosed separate data extraction and subsequent analysis processes in a cloud environment. Laribi taught a performance monitoring service for cloud-based applications that adaptively provisions or de-provisions application instances based on load.
  • Motivation to Combine: A POSITA would combine Laribi’s load-based scaling techniques with Safford’s system to efficiently manage computational resources. Since social media message volume is variable, applying Laribi’s method to monitor the load on Safford’s analysis process (i.e., the extraction rate) and dynamically adjust the number of parallel processors would have been an obvious way to ensure timely analysis while conserving resources.
  • Expectation of Success: A POSITA would have a high expectation of success, as both Safford and Laribi described systems operating in cloud-based environments designed for scalable data processing.

Ground 6: Obviousness over Safford, Chen, and Masters - Claims 1-2, 8-12, 15-16, and 19-20 are obvious over Safford, Chen, and Masters.

• Prior Art Relied Upon: Safford (Patent 9,288,123), Chen (Patent 8,010,524), and Masters (Patent 7,051,098).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground presented an alternative to Laribi for the dynamic scaling limitations. Petitioner argued that Masters taught a system for managing resources for interrelated and dependent applications. Masters’s Quality of Service (QoS) managers recommend scaling applications up or down to maintain performance, and its Resource Manager determines where to place new application copies.
  • Motivation to Combine: A POSITA would find Masters’s teachings on managing dependent applications directly pertinent to Safford’s dependent data extraction and subsequent analysis processes. To handle fluctuating data rates from social media and ensure timely brand-related issue response, a POSITA would apply Masters’s resource allocation principles to scale Safford’s analysis processors based on the rate of incoming extracted data.
  • Expectation of Success: The combination involved applying known resource management techniques (Masters) to a known data processing workflow (Safford), representing a predictable implementation for a POSITA.

Ground 11: Obviousness over Heath, Randall, and Laribi - Claims 1-2, 8-12, 15-16, and 19-20 are obvious over Heath, Randall, and Laribi.

• Prior Art Relied Upon: Heath (Patent 8,909,771), Randall (Patent 7,725,452), and Laribi (Patent 9,769,085).

• Core Argument for this Ground:

  • Prior Art Mapping: This ground used Heath as the primary reference. Petitioner asserted that Heath taught a “forum analyzer” that scours the internet for consumer feedback based on keywords, calculates relevance scores to select data sources, and monitors those sources. Randall was cited for its disclosure of crawlers that detect online content changes on a set schedule (e.g., every few minutes). The combination of Heath and Randall was argued to teach the scheduled monitoring and extraction of data from selected sources. Laribi was then combined for the same purpose as in Ground 1: to provide dynamic, load-based resource scaling.
  • Motivation to Combine: A POSITA would combine Randall’s scheduled crawling with Heath’s forum analysis to create a system for regular, automated monitoring. A POSITA would then integrate Laribi’s adaptive provisioning to allow the combined Heath/Randall system to efficiently scale its analysis capacity in response to the volume of extracted consumer feedback, ensuring timely sentiment analysis.
  • Expectation of Success: Integrating these known technologies—data scouring (Heath), scheduled crawling (Randall), and load-based scaling (Laribi)—was argued to be a straightforward design choice for a POSITA seeking to build an efficient, scalable social media monitoring platform.

• Additional Grounds: Petitioner asserted numerous additional obviousness challenges, primarily by adding single-purpose references to the core combinations above. These included combinations with Seibel (for sourcing data from open and closed-source databases), Schultz and Gravano (for query expansion and using accuracy thresholds), and Moulinier (for natural language query translation).

4. Key Claim Construction Positions

• Petitioner contended that certain claim terms, while indefinite for purposes of infringement in parallel litigation, do not render a patentability determination impossible in an IPR. Petitioner argued that the core meanings of the terms are understood by a POSITA and are met by the prior art.
• "extraction rate" / "first processing rate": Petitioner asserted these terms cover determining a quantity of data output or analyzed over a period of time.
• "[to] dynamically adjust the number of parallel processors": Petitioner asserted this term covers adjusting distinct processors operating in parallel, arguing that the prior art’s teachings on adaptive provisioning based on load or performance metrics meet this limitation.

5. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial under Fintiv is inappropriate.
• Petitioner stipulated that, should the Board institute review, it will not pursue in the parallel district court litigation the same grounds raised in the petition or any grounds that could have reasonably been raised.
• Petitioner also argued that the trial date in the parallel litigation is highly uncertain, as the case is one of twelve scheduled for jury trial selection on the same day before the same judge, making it unlikely to proceed as scheduled.

6. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-20 of the ’679 patent as unpatentable under 35 U.S.C. §103.