Searchplex
  • Case Studies
  • Blog
  • About Us
Search Stack Audit
Search Stack Audit

Browse services, solutions, case studies, and Vespa consulting pages from the same navigation tree as desktop.

Case StudiesBlogAbout Us
Searchplex

AI-native search and retrieval engineering for enterprises where search drives revenue, productivity, and customer experience.

GitHubLinkedInTwitterYouTube
Amsterdam, Netherlands

Industries

  • News & Media
  • Finance Technology
  • Legal Technology

Quick Links

  • Diagnostic
  • Vespa.ai Consulting
  • AI Search & Personalization
  • Enterprise RAG
  • Visual Search
  • Retrieval architecture for AI
  • Case Studies
  • Blog

Company

  • About Us
  • Events
  • Subscribe
  • Contact
  • hello@searchplex.net
© 2026 Searchplex. All rights reserved.
Cookie SettingsPrivacy PolicyTerms of Service
    1. Home/
    2. Case Studies/
    3. IPRally
    Intellectual Property Technology

    IPRally: Vespa Architecture Audit for a Production Patent Search Platform

    Searchplex worked with IPRally to audit a production patent search platform built on Vespa, reviewing architectural decisions across search behavior, data representation, infrastructure, and performance optimization.

    Intellectual Property Technology·Vespa Review & Optimization
    IPRally logo
    IPRally

    IPRally builds what it describes as the most accurate patent search engine on the planet. Their SaaS platform serves patent professionals, IP researchers, and competitive intelligence teams who depend on precise retrieval across large global patent datasets.

    Industry
    Intellectual Property Technology
    Hosting
    Google Cloud Platform
    Use Cases
    Patent search, Vespa architecture audit, performance optimization
    Solution
    Vespa.ai
    Engagement
    Vespa Review & Optimization

    IPRally: Vespa Architecture Audit for a Production Patent Search Platform

    As search systems mature, the hardest questions are no longer just about relevance. They are about whether query execution, infrastructure, and data modeling are making the right trade-offs for performance and scale.

    Searchplex worked with IPRally to audit a production patent search platform built on Vespa, reviewing architectural decisions across search behavior, data representation, infrastructure, and performance optimization.


    Client

    IPRally builds what it describes as “the most accurate patent search engine on the planet.” Their SaaS platform serves patent professionals, IP researchers, and competitive intelligence teams who depend on precise retrieval across large global patent datasets.

    The platform is built on Vespa and operates as a production system serving real client workloads.


    The situation

    IPRally had already built a sophisticated patent search platform on Vespa and the system was working well in production. The engineering team had made deliberate architectural decisions and the platform was successfully serving customers.

    As the company prepared for the next phase of growth, the CTO wanted an independent expert review before scaling further.

    The goal of the engagement was not to fix a broken system, but to answer a different set of questions:

    • Are we using Vespa in the most effective way?
    • Are there implementation patterns that could become expensive or difficult to maintain at higher scale?
    • Are there capabilities in Vespa that we are not yet using that could improve performance or simplify the system?

    Searchplex was engaged to conduct a structured architecture audit of the Vespa deployment and provide recommendations across several areas of the search stack.

    The review focused on key parts of the system including:

    • custom search behavior implemented in the Vespa application
    • wildcard search functionality
    • patent data representation and storage
    • infrastructure and deployment choices
    • debugging and performance analysis practices

    This pattern is common in mature search systems. Once a product is working successfully in production, the next level of improvement often comes from validating architectural trade-offs and identifying the highest-impact optimization opportunities.


    What we reviewed

    Searchplex conducted a structured audit of IPRally’s Vespa-based patent search system hosted on Google Cloud Platform.

    The review examined how the platform used Vespa across several core areas.

    Custom search logic

    We evaluated how custom search logic was implemented within the Vespa application and where the current approach aligned well with Vespa best practices versus where complexity could potentially be reduced.

    Wildcard search behavior

    Wildcard queries can be computationally expensive and difficult to optimize cleanly. The review examined how these queries were handled and identified opportunities to improve their performance characteristics.

    Patent data representation

    Patent search systems have unique data modeling requirements. We reviewed how patent data was represented and indexed within Vespa to identify opportunities for improving efficiency and reducing processing overhead.

    Infrastructure and deployment

    The system infrastructure was reviewed to assess trade-offs between different deployment approaches, with the goal of supporting long-term scalability and operational efficiency.

    Debugging and performance analysis

    We assessed the existing methods used to investigate search behavior and performance issues and provided recommendations for improving observability and troubleshooting.

    The engagement produced both immediate suggestions and longer-term architectural guidance aimed at improving performance, scalability, and maintainability.


    Architecture takeaway

    This engagement reflects a pattern frequently seen in mature search platforms:

    • Production search systems can still contain hidden inefficiencies in query execution, data modeling, or infrastructure configuration.
    • Architecture audits are most valuable after initial product success, when teams want to validate trade-offs rather than simply ship new features.
    • Patent search systems place pressure on both retrieval quality and system design, making efficiency, observability, and architectural clarity especially important.
    • Independent expert reviews help teams confirm which design decisions are sound and where targeted improvements will have the greatest long-term impact.

    Outcome

    The audit delivered a comprehensive written review of the Vespa implementation along with a prioritized roadmap for future optimization.

    Some of the expected benefits identified through the review included:

    • reduced wildcard query latency
    • lower infrastructure resource utilization
    • elimination of inefficient post-filtering processes

    At the time of the review, these recommendations had not yet been implemented. The primary outcome was architectural validation — confirming that the core system design was sound while clarifying the trade-offs involved in future improvements.

    The optimization roadmap provides IPRally with a structured path to implement performance and efficiency improvements as the system continues to scale.

    "

    Searchplex delivered a thorough and well-structured analysis within the agreed timeline. The report is comprehensive and addresses all the areas defined in the scope of work. Ravindra communicated well with us, was receptive to our feedback, and provided good suggestions throughout the work.

    "
    CTO, IPRally
    Engagement summary
    Industry

    Patent Search / Intellectual Property Technology

    Starting platform

    Production patent search system built on Vespa

    Hosting
    Google Cloud Platform
    Engagement type

    Vespa Review & Optimization (Search Stack Audit)

    Focus areas

    Custom search behavior, wildcard search, patent data representation, infrastructure, debugging and performance

    Primary value delivered

    Architectural validation and prioritized optimization roadmap

    Expected improvement areas

    Query latency, infrastructure efficiency, streamlined search processing

    Searchplex conducted an expert audit of IPRally’s Vespa-based search platform to validate architectural decisions, uncover inefficiencies, and deliver a prioritized roadmap for improving performance, scalability, and operational efficiency.


    Related

    Vespa Review & Optimization →
    Search Stack Audit →

    In this article

    1. IPRally: Vespa Architecture Audit for a Production Patent Search Platform
    2. Client
    3. The situation
    4. What we reviewed
      1. Custom search logic
      2. Wildcard search behavior
      3. Patent data representation
      4. Infrastructure and deployment
      5. Debugging and performance analysis
    5. Architecture takeaway
    6. Outcome
    7. Related
    Intellectual Property Technology · Retrieval engineering

    Need similar retrieval work in production?

    If your team is evaluating retrieval architecture, migration, or AI search work like this, the next step is usually a direct discussion about the system, constraints, and delivery path.

    Contact Searchplex
    More case studies
    Tax & Regulatory Technology

    Nextens

    Nextens, part of LexisNexis Risk Solutions & RELX, is a leading Dutch provider of tax solutions and digital products for fiscal and financial professionals, used daily by accountants, tax advisors, and finance teams.
    Read case study →
    Legal Technology

    CuratedAI

    CuratedAI is a Belgium-based legal technology company. Their platform makes EU legislation and case law searchable in natural language - down to the specific article or paragraph - for legal professionals working in privacy and IT law.
    Read case study →