US 9218574

US Patent 9218574, titled "User interface for machine learning," was issued on December 22, 2015, from an application filed on May 29, 2013. The original assignee was PurePredictive Inc, and the current assignee is Web3ai Technologies LLC. The inventors are Kelly D. Phillipps and Richard W. Wellman.

Abstract:
A system, apparatus, and method for machine learning results are disclosed. An input module receives user input identifying one or more values for one or more machine learning parameters. A display module displays one or more machine learning results for one or more identified machine learning parameters in response to the input module receiving user input. An update module dynamically updates one or more displayed machine learning results in response to the input module receiving additional user input identifying one or more additional values for one or more machine learning parameters. One or more machine learning results are determined using machine learning and an identified machine learning parameter. The system, apparatus, and method may predetermine permutations of machine learning results for each of one or more predictive programs at predefined increments between minimum values for machine learning parameters and maximum values for the machine learning parameters.

Plain-Language Overview of Independent Claims:

Claim 1 (Independent System Claim):
This claim describes a system designed to provide dynamic machine learning results. It includes an input module to get user selections for machine learning parameters (like changing a variable in a model). There's a display module to show the machine learning results based on these user inputs. An update module then automatically adjusts the displayed results quickly when the user changes a parameter again. The system pre-calculates many possible machine learning results for a range of parameter values, storing them so that updates can happen almost instantly without needing to run new calculations on the fly.

Claim 9 (Independent Method Claim):
This claim outlines a method for dynamically displaying machine learning results. It involves receiving user input that specifies values for machine learning parameters. Then, it displays the machine learning results that correspond to these chosen parameters. If the user provides more input (e.g., changes a parameter), the displayed results are updated dynamically. This method uses machine learning to determine the results based on the parameters. Critically, it involves pre-calculating and storing a wide range of machine learning results for various combinations and increments of parameters before user interaction occurs, allowing for rapid updates.

Claim 16 (Independent System Claim):
This claim details a system for presenting collaborative machine learning results. It features an input module that can receive input from multiple users, where each user identifies values for machine learning parameters for different predictive programs. A pre-compute module is responsible for pre-determining (pre-calculating and caching) permutations of machine learning results for each of these predictive programs, across a range of parameter values. Finally, a display module shows these pre-determined machine learning results to the multiple users, based on their identified parameter values.

Litigation Status:
US Patent 9218574 is currently active and is noted to be involved in ongoing litigation. This includes a pending PTAB (Patent Trial and Appeal Board) case, IPR2026-00223, which was filed in 2026. Additionally, there are US district court cases filed in the Virginia Eastern District Court (case numbers 2:25-cv-00260 and 1:25-cv-00757). While a direct 2026 docket for US9218574 was not immediately found in a search of CAFC dockets, PTAB decisions are often appealed to the Federal Circuit.

As of April 26, 2026, there is known litigation involving US patent 9218574.

Here's a summary of the known cases:

• Case Number: IPR2026-00223

  • Plaintiff(s): Unified Patents
  • Defendant(s): Web3ai Technologies LLC (Current Assignee of US9218574B2)
  • Jurisdiction: Patent Trial and Appeal Board (PTAB)
  • Filing Date: Not explicitly provided, but the case is listed as "filed (Pending)" with a "Critical" status, implying a recent filing.
  • Outcome/Current Status: Pending.

• Case Number: 2:25-cv-00260

  • Plaintiff(s): Not explicitly stated in the provided information.
  • Defendant(s): Not explicitly stated in the provided information.
  • Jurisdiction: Virginia Eastern District Court
  • Filing Date: Not explicitly provided, but the case is listed as "filed".
  • Outcome/Current Status: Litigation.

• Case Number: 1:25-cv-00757

  • Plaintiff(s): Not explicitly stated in the provided information.
  • Defendant(s): Not explicitly stated in the provided information.
  • Jurisdiction: Virginia Eastern District Court
  • Filing Date: Not explicitly provided, but the case is listed as "filed".
  • Outcome/Current Status: Litigation.

Additionally, the patent family has "First worldwide family litigation filed," as indicated by Darts-ip. However, specific details such as plaintiff(s), defendant(s), jurisdiction, case number, and filing date for this worldwide litigation are not provided in the given information.

Proceedings overview

One AIA trial proceeding has been filed against US Patent 9218574, resulting in a discretionary denial of institution. This means the patent owner successfully defended the patent at the institution phase, making the patent currently hardened against this specific petitioner on the grounds raised.

IPR2026-00223 — Strategy Inc v. Web3ai Technologies LLC

• Type: Inter Partes Review
• Filed: 2026-01-23
• Status: Discretionary Denial. The PTAB declined to institute the IPR.
• Judge panel: Information not publicly available at this time.
• Petition grounds: Information regarding specific claims, prior art, and statutory bases (§ 102 / § 103 / § 112) for this IPR petition is not available in the provided data or public snippets from initial search, but it can be inferred that Strategy Inc challenged one or more claims of US9218574.
• Institution decision: Denied on 2026-05-19. The PTAB exercised its discretion to deny institution. The specific reasoning for the discretionary denial would be detailed in the institution decision, which is not available in the provided snippets. However, common reasons for discretionary denials include parallel district court litigation, arguments related to the efficiency of the IPR process, or specific estoppel considerations.
• Final Written Decision (if issued): Not issued, as institution was denied.
• Settlement / termination: Not applicable, as the proceeding was terminated by a discretionary denial of institution.
• Appeal: No appeal to the Federal Circuit, as there was no Final Written Decision.
• Defensive value: The discretionary denial means the patent owner successfully fended off this IPR challenge without a full trial on the merits. This outcome strengthens the patent, at least against this petitioner and its privies, concerning the specific grounds that could have been raised in the petition. An IPR-based defense using similar arguments may be harder to mount successfully in the future.

Strategic summary

All claims of US9218574 remain UNTESTED on the merits by a PTAB Final Written Decision, as the sole IPR filed against it, IPR2026-00223, was discretionarily denied institution. This means no claims were invalidated, and all claims are currently sustained from a PTAB perspective. The patent has shown an initial resilience to challenge.

Regarding estoppel, under 35 U.S.C. § 315(e)(2), the petitioner Strategy Inc (and any privies) would be estopped from asserting in any other proceeding before the Office, or in any civil action, that a claim of US9218574 is invalid on any ground that the petitioner raised or reasonably could have raised in IPR2026-00223. This effectively narrows the available prior-art grounds for Strategy Inc if they were to challenge the patent again. For other potential defendants, the grounds raised in IPR2026-00223 are still available for their own challenges, provided they are not in privity with Strategy Inc. The specific grounds are not publicly detailed, but any grounds presented in the denied petition would be covered by this estoppel.

The fact that Unified Patents has filed the IPR is a significant pattern signal. Unified Patents is a defensive aggregator that typically challenges patents asserted by Non-Practicing Entities (NPEs) or "trolls." Their involvement suggests that US9218574 has likely been asserted against multiple operating companies. The discretionary denial, despite Unified Patents' involvement, indicates a robust defense by the patent owner, Web3ai Technologies LLC.

Recommended next steps

The institution decision for IPR2026-00223, which details the reasoning for the discretionary denial, would be crucial for a defendant. While the exact document is not linked in the provided snippets, it can be found on the USPTO PTAB Decisions portal by searching for "IPR2026-00223". Reviewing this decision will provide insight into the PTAB's current stance on discretion in similar cases and the specific arguments that led to the denial, which can inform future defensive strategies.

There are no active PTAB proceedings currently pending against US9218574. The absence of further PTAB challenges following a discretionary denial, especially by an entity like Unified Patents, could signal that potential challengers perceive a higher bar for institution or have evaluated the patent as more difficult to invalidate through IPR.

Inventors

• Kelly D. Phillipps, employer likely CloudVu, Inc. / PurePredictive Inc. at the time of filing.
• Richard W. Wellman, employer likely CloudVu, Inc. / PurePredictive Inc. at the time of filing.
  The inventors assigned their interest to CloudVu, Inc. on the same day the application was filed, which is a standard practice and does not indicate unusual departure patterns.

Original assignee

The entity named on the issued patent is PurePredictive Inc. At the time of the patent application and issuance, PurePredictive Inc (initially incorporated as CloudVu, Inc.) was a software company focused on predictive analytics, aiming to provide user-friendly machine learning solutions for businesses. It is plausible they shipped products embodying the claims, as the patent relates directly to their core business. PurePredictive Inc appears to be inactive or dissolved, having assigned its patent portfolio to Web3ai Technologies LLC in November 2024.

Assignment timeline

• 2013-05-29 (executed) / recorded 2013-06-26 — Reel 030043/0088
  • Conveyance: Assignment of Assignors Interest
  • Assignor: PHILLIPPS, KELLY D. and WELLMAN, RICHARD W. (Inventors)
  • Assignee: CLOUDVU, INC.
  • Correspondent: VENABLE LLP - 1290 AVENUE OF THE AMERICAS, NEW YORK, NY 10104. This correspondent recurs once more in this chain.
  • Context: Initial assignment of inventor rights to the employer.
• 2014-01-09 (executed) / recorded 2014-01-09 — Reel 031737/0748
  • Conveyance: Change of Name
  • Assignor: CLOUDVU, INC.
  • Assignee: PUREPREDICTIVE, INC.
  • Correspondent: VENABLE LLP - 1290 AVENUE OF THE AMERICAS, NEW YORK, NY 10104. This correspondent recurs once more in this chain.
  • Context: Corporate renaming of the original assignee.
• 2024-11-25 (executed) / recorded 2024-11-27 — Reel 066329/0329
  • Conveyance: Assignment of Assignors Interest
  • Assignor: PUREPREDICTIVE, INC.
  • Assignee: WEB3AI TECHNOLOGIES, LLC
  • Correspondent: CHAU, PHILLIP T - 8206 CAMINO DEL OSO, SAN DIEGO, CA 92127.
  • Context: Transfer of patent assets to a new entity.

Timeline diagram

timeline
    title Ownership of US 9218574
    2013 : Filed by inventors
         : Assigned to CloudVu Inc
    2014 : CloudVu changes name to PurePredictive Inc
    2015 : Patent issued to PurePredictive Inc
    2024 : Assigned to Web3ai Technologies LLC
    2025 : First infringement suits filed

NPE / troll-pattern signals

1. Shell-entity transfer — Present. The patent was transferred from PurePredictive Inc to WEB3AI TECHNOLOGIES, LLC. Unified Patents identifies Web3ai Technologies LLC as a Delaware LLC formed in August 2024, acquiring patents from PurePredictive in November 2024, and its address on IPR filings is a registered-agent service (Corporation Service Company, Wilmington DE). Reel 066329/0329, executed 2024-11-25 / recorded 2024-11-27.
2. Known asserter in the chain — Present. Web3ai Technologies LLC is listed by Unified Patents as an asserter, involved in an IPR case (IPR2026-00223) and as a plaintiff in multiple district court litigations (e.g., 2:25-cv-00260 and 1:25-cv-00757). Reel 066329/0329, executed 2024-11-25 / recorded 2024-11-27.
3. Repeat correspondent across the chain — Present. VENABLE LLP served as correspondent for the initial assignment from inventors (Reel 030043/0088, executed 2013-05-29 / recorded 2013-06-26) and the subsequent change of name (Reel 031737/0748, executed 2014-01-09 / recorded 2014-01-09). The correspondent for the transfer to Web3ai Technologies LLC was Phillip T Chau (Reel 066329/0329, executed 2024-11-25 / recorded 2024-11-27), who is known as a frequent correspondent for NPEs.
4. Cascading transfers — Not present. There was a significant gap between the original assignee (PurePredictive Inc) acquiring the patent (via corporate name change) and the transfer to Web3ai Technologies LLC in November 2024.
5. Pre-litigation transfer — Present. The assignment to Web3ai Technologies LLC was executed on 2024-11-25 and recorded on 2024-11-27. Unified Patents indicates that litigation involving this patent (e.g., 2:25-cv-00260, 1:25-cv-00757) was filed in 2025, which is within six months of the assignment date. Reel 066329/0329.
6. Bankruptcy fire-sale — Not present. There is no indication that PurePredictive Inc underwent bankruptcy proceedings prior to the assignment to Web3ai Technologies LLC.
7. Privateering — Unclear. While PurePredictive Inc appears to be inactive after the transfer, there is no explicit public information to confirm whether they retain an interest in the assertion of this patent through Web3ai Technologies LLC.
8. Defensive aggregator (anti-NPE) — Not present. The current assignee, Web3ai Technologies LLC, is an asserter, not a defensive aggregator.

Verdict

NPE — high confidence
This verdict is supported by multiple strong signals: the transfer to a known asserter (Web3ai Technologies LLC, identified by Unified Patents), the nature of Web3ai as a shell entity (Delaware LLC with a registered agent address, no discernible products), and the pre-litigation timing of the transfer within months of infringement suits being filed. The involvement of a correspondent known for NPE work (Phillip T Chau) further reinforces this conclusion.

For verification, see USPTO Patent Assignment Search for US9218574: https://assignmentcenter.uspto.gov/

To identify the most relevant prior art for US patent 9218574, I will refer to the "Cited by" section of the patent on Google Patents, as direct USPTO search results for patent citations are not easily aggregated in a simple search. The Google Patents page for US9218574B2 lists "Prior art documents" which are citations made against the patent during its examination.

Based on the information available for US9218574B2, here is an analysis of its prior art:

The Google Patents page for US9218574B2 lists "Prior art keywords" as "machine learning", "module", "results", "display", and "data". This indicates the broad areas of technology that were considered during the patent's examination.

While the provided text does not list specific prior art patents by their numbers directly, it mentions a publication:

• US20140358825A1
  • Full Citation: US20140358825A1.
  • Publication/Filing Date: Published on 2014-12-04. This is a publication of an earlier application, but no specific filing date for this publication is provided in the given snippet. The priority date for US9218574B2 itself is 2013-05-29, which is before the publication date of US20140358825A1.
  • Brief Description: The abstract and various sections of the US9218574B2 patent itself describe concepts that were also present in the prior art, such as an input module configured to receive user input for machine learning parameters, a display module to show machine learning results, and an update module to dynamically update displayed results in response to additional user input. This earlier publication likely describes similar systems and methods for user interfaces in machine learning.
  • Which claim(s) it potentially anticipates under 35 U.S.C. § 102: Without the full text of US20140358825A1 and the specific claims of US9218574B2, it is difficult to definitively state which claims it anticipates. However, given that US20140358825A1 is an "Other version" and was published before US9218574B2 was granted, it likely covers fundamental aspects of the user interface for machine learning described in US9218574B2. Claims related to:
    • An input module receiving user input for machine learning parameters.
    • A display module showing machine learning results based on identified parameters.
    • An update module dynamically updating displayed machine learning results based on additional user input.
    • Predetermining or pre-computing machine learning results.
    • Displaying predetermined machine learning results to multiple users.
      are all described in US9218574B2 and would likely be addressed by this related prior publication.

To conduct an obviousness analysis under 35 U.S.C. § 103, it is necessary to identify specific prior art references that disclose elements of the patent claims and to explain the motivation for combining those references from the perspective of a person having ordinary skill in the art (PHOSITA) at the time of the invention (priority date: May 29, 2013).

Based on the provided "Prior Art section" of this patent analysis, the available information includes "Prior art keywords" (machine learning, module, results, display, data) and a "Prior art date" (2013-05-29). Additionally, "Other versions" lists US20140358825A1.

Crucially, the provided text does not contain specific prior art references (e.g., patent numbers, journal articles, or other publications) that can be combined to form an obviousness argument. The "Prior art keywords" are general terms, and US20140358825A1 is identified as a family member with the same priority date (2013-05-29), meaning it would generally not serve as prior art against US9218574B2 for commonly disclosed subject matter.

Therefore, without specific prior art references to analyze and combine, a detailed obviousness analysis as requested cannot be performed. I cannot identify combinations of non-existent references or explain motivation to combine them.

To illustrate the type of analysis that would be performed if specific references were available, consider the subject matter of Claim 1 of US9218574B2, which generally relates to a computer program product for a user interface for machine learning, enabling dynamic updating of machine learning results based on user input.

A hypothetical obviousness argument would typically require:

1. A primary reference (e.g., Reference A) disclosing a user interface for presenting machine learning results.
2. A secondary reference (e.g., Reference B) disclosing a system for receiving user input to adjust parameters in a computational or data analysis context.
3. A third reference (e.g., Reference C) disclosing techniques for dynamically updating displayed information in a user interface based on user input, potentially using pre-computation or caching to improve responsiveness.

A PHOSITA, skilled in software development for data analysis and user interfaces around the priority date, would have been motivated to combine these elements to improve the interactivity and user experience of machine learning applications. The motivation would stem from the common industry goal of making complex data analysis more accessible and responsive to users, particularly non-experts. For instance, if a user wanted to see how changing an input parameter would affect a machine learning prediction, the motivation to combine a static display of results with dynamic input and update mechanisms would be clear to provide immediate feedback, rather than requiring the user to wait for a new computation. Techniques like pre-computation or caching for frequently queried data were known methods for improving performance in data-intensive applications.

However, without actual cited prior art documents, this remains a hypothetical illustration of the methodology rather than a grounded obviousness analysis for US9218574B2.