Plaid Technologies Inc v. Yodlee Inc

1. Case Identification

• Case #: CBM2016-00088
• Patent #: 7,752,535
• Filed: June 10, 2016
• Petitioner(s): Plaid Technologies, Inc.
• Patent Owner(s): Yodlee, Inc.
• Challenged Claims: 1-10

2. Patent Overview

• Title: Categorization of Summarized Information
• Brief Description: The ’535 patent relates to systems and methods for automatically collecting, categorizing, and summarizing transaction information from third-party internet websites. The patent further discloses using stored historical transaction information to predict future transaction statistics.

3. Grounds for Unpatentability

Ground 1: Ineligibility Under §101 - Claims 1-10 are directed to an abstract idea

• Core Argument:
  • Petitioner argued that claims 1-10 are directed to the patent-ineligible abstract idea of categorizing and summarizing past financial transactions and using that information to predict future transactions. Petitioner asserted this is a longstanding, fundamental economic practice that is not transformed into a patentable invention merely by implementing it on a generic computer.
  • Under the Alice two-step framework, Petitioner contended the claims fail both steps. For Step 1, the claims recite the core abstract idea without limiting it to a specific, inventive application. For Step 2, the claims add no inventive concept, instead reciting only a collection of generic and conventional components and steps. These include a "computer system," "proprietary software," and routine functions like collecting data from websites, storing it, categorizing it based on pre-set rules, and reporting summaries.
  • Petitioner further argued that dependent claims 2-5 and 7-10 fail to add an inventive concept, as they merely introduce conventional post-solution activities. For example, claims 5 and 10, which require reporting transaction summaries "through the Internet network," do not save the claims because using the internet for data transmission is a well-understood, routine activity.

Ground 2: Obviousness over Schutzer, Knoblock, and Lazarus - Claims 1, 5, 6, and 10 are obvious over the combination

• Prior Art Relied Upon: Schutzer (Patent 5,920,848), Knoblock ("Modeling Web Sources for Information Integration"), and Lazarus (Patent 6,430,539).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that the combination of Schutzer, Knoblock, and Lazarus taught every limitation of the challenged claims.
    • Schutzer was argued to disclose an integrated financial system using "intelligent agents" to collect, categorize based on user-defined rules, store, and report financial transaction data from multiple bank servers over a network. It taught a graphical user interface for users to request transaction summaries over specific date ranges and by categories like payee.
    • Knoblock was cited to supply the teaching of automatically navigating to various third-party web sources, parsing semi-structured HTML pages, and extracting information using "wrappers" or site-specific scripts to create a unified view of the gathered data.
    • Lazarus was argued to teach the final key elements: analyzing historical consumer financial behavior to predict future spending. It disclosed using probabilistic algorithms to create and update "description characteristics" (termed "merchant vectors") for categorizing transactions and using this analysis to forecast future statistical information.
  • Motivation to Combine: A person of ordinary skill in the art (POSITA) would have been motivated to improve Schutzer's financial aggregation system using the known techniques from Knoblock and Lazarus. A POSITA would combine Schutzer with Knoblock to automate the data collection process and expand its reach beyond direct bank connections to the broader web, a predictable improvement. A POSITA would then incorporate Lazarus's predictive analytics and automated categorization methods to enhance the Schutzer/Knoblock system, providing more advanced and commercially valuable financial insights, which was the goal of such systems.
  • Expectation of Success: The combination involved applying known web data extraction techniques (Knoblock) and established predictive financial modeling (Lazarus) to a conventional financial data aggregation platform (Schutzer). Each reference would have performed its intended function, leading to the predictable result of a more automated and intelligent financial management tool with a high expectation of success.

4. Key Claim Construction Positions

• Term: "a collection function automatically navigating to and retrieving transaction information... from third-party Internet-connected web sites and gathering information concerning transactions" (claims 1 and 6)
  • Petitioner proposed this should be construed as "[s]oftware that collects a natural person's transaction information from two or more websites, where for each website it parses a web page rendered by a browser..." This construction was argued to be critical for mapping Knoblock's teachings of automated web navigation and HTML parsing onto the claims.
• Term: "a definition of purpose of transaction including at least expenditure types" (claims 1 and 6)
  • Petitioner proposed this should be construed as "a definition of a payee, account or other identifier associated with a transaction." This broad construction was used to argue that the user-defined expense categories and query parameters disclosed in Schutzer met this limitation.

5. Relief Requested

• Petitioner requested institution of a Covered Business Method (CBM) review, a judgment that claims 1-10 are unpatentable under 35 U.S.C. § 101 and claims 1, 5, 6, and 10 are invalid as obvious under 35 U.S.C. § 103, and cancellation of all challenged claims.