Welcome to SynapseIP, an advanced data and analytics platform directed to artificial intelligence (AI) and machine learning (ML) intellectual property (IP). This help center includes documentation describing the platform's various features and terminology, as well as guides to the user interfaces and workflows on the platform.
SynapseIP is an IP platform specific to AI/ML data and analytics. The platform combines hybrid semantic search, trend analysis, and IP overview information to provide an integrated and in-depth understanding of the IP landscape as it relates to AI/ML innovations and investments and the entities active in this space. The platform is built on a relational database system that includes 56,000+ AI/ML-related patents and publications dating back to 2023.
Each entry (i.e., patent or publication) in the database is enriched with metadata and context; specifically, each entry corresponds to a patent or publication with multiple associated AI embeddings. The multiple AI embeddings enable accurate and robust semantic searching over multiple fields and combinations of fields.
Metadata and context for each entry further include the assignee name (i.e., owner). The SynapseIP platform normalizes each assignee name to ensure that the AI/ML IP assets held by different entities are accurately and comprehensively represented.
SynapseIP is designed with a streamlined user interface divided between three primary web service pages:
Hybrid search, trend visualization, and proactive alerts
The Search & Trends page is the primary interface for discovering and monitoring granted patents and published applications. It combines powerful search capabilities with visual trend analytics to provide an easily comprehensible view of the AI/ML IP landscape. The intuitive interface allows users to construct complex queries using both keywords and semantic similarity, filter results by various metadata fields, and visualize filing trends over time, by CPC classification, or by assignee.
The Search & Trends page further includes the option to save a particular search configuration as an alert. SynapseIP updates its database on a weekly basis, following the USPTO schedule. Saved searches are automatically run when new data becomes available, and users are notified of new matches.
Example Use Cases: Ongoing competitive monitoring, prior art searches, freedom-to-operate and clearance analysis, and staying current with AI/ML IP as it relates to specific technology areas. Graphs provide visual guides on trends across the AI/ML IP landscape.
Insights on how busy specific technology areas are and where opportunities may exist
The IP Overview page provides a quantitative view of patent and publication activity within a defined scope. This page includes analysis and insights for saturation, activity rate, and momentum measurements, as well as identifying CPC codes under which patent and publications are concentrated. The interface presents four primary metrics (subject matter saturation, patent and publication activity rate, patent grant and publication momentum, and top CPC codes) supported by a monthly trend line, CPC distribution chart, and sortable results table with direct links to the underlying patents and publications.
An optional “Group by Assignee” toggle enables the KNN Sigma graph visualization with confidence-scored signals highlighting potential gaps, bridging opportunities*, focus convergence, and crowd-out patterns.
Example Use Cases: AI/ML investment decisions, R&D opportunity identification, competitive threat assessment, and understanding where R&D focus is shifting in and around specific technology areas in the context of AI/ML.
Claim-level semantic comparison for rapid FTO and infringement triage
The Scope Analysis page supplements AI/ML IP search, trends, and IP overview by providing freedom-to-operate (FTO) and infringement-risk screening. Input a natural language description of subject matter of interest (e.g., product features, invention disclosures, draft claims, etc.) and run a semantically comparison against independent claims across patents in the SynapseIP database. Closest matches with context-rich analysis is returned.
Results are visualized in an interactive node map that positions the user input at the center with claim nodes arranged radially by similarity, while a synchronized evidence table lists the associated patents, assignees, publication dates, and full claim language. This combination provides an immediate, data-driven starting point for preliminary freedom-to-operate, infringement-risk reviews, and design-around brainstorming.
Example Use Cases: pre-FTO review, infringement-risk screening, due diligence, clearance analysis.
SynapseIP uses Auth0 for secure authentication. Log in or sign up using the button in the top navigation bar. All features require authentication to ensure data security and usage tracking.
Search & Trends provides direct and easy access to the AI/ML patent and publication data set available through SynapseIP. Semantic queries, keyword input, and CPC filters are available to narrow results. Trend groupings may assist in understanding filing patterns.
Relevant or important searches can be saved as alerts to avoid repeated manual runs. With search criteria of interest input, click 'Save as Alert' to receive notifications when new patents or publications match that criteria. Alerts can be managed through the navigation bar (hover over username for menu).
Input a product description, invention disclosure, or draft claims into the Scope Analysis page to generate a semantic comparison against independent claims. Use the similarity graph and evidence table to identify relevant patents to assess infringement risks.
Navigate to the IP Overview page to view how busy a technology area is. Enter focus keywords and/or CPC codes, review the saturation/activity rate/momentum tiles, inspect the timeline and CPC bars, and review the result set table for representative patents and publications.
Use CSV/PDF exports for shareable reports, and copy links from the IP Overview result set table to access full text and figures of specific patents or publications.
A “bridging” patent/publication is one in which the invention is directed to one technology area but the scope of protection can be broadened to cover other areas. Example: a patent claiming an improvement to internal combustion engines in automobiles, and the improvement can also be used in aviation, marine, and other internal combustion engine applications. Bridging patents have been shown to be especially commercially valuable. See, e.g., Choi & Yoon, Measuring Knowledge Exploration Distance at the Patent Level, 16 J. Informetr. 101286 (2002) (linked here). See also Moehrle & Frischkorn, Bridge Strongly or Focus — An Analysis of Bridging Patents [...], 15 J. Informetr. 101138 (2001).
CPC (Cooperative Patent Classification) codes categorize patents and applications by technology area. For reference, the USPTO generally assigns AI and machine learning subject matter under one of the following CPC section (letter)+class(number)+subclass(letter) classifications: A61B, B60W, G05D, G06N, G06V, and G10L. More specific AI/ML-related subject matter is generally assigned to group, as well, as indicating by a number appended to the subclass: G06F17, G06F18, G06F40, G06K9, G06T7. A further subgroup indicates subject matter at an even more granular level, which is indicated by a third number, preceded by a backslash. For AI/ML-related subject matter, this is most commonly encoutered in CPC classification G06F16/90.
Review the SynapseIP platform Terms of Service, Privacy Policy, and Data Processing Agreement.
If you have questions or encounter issues not covered in this documentation, please contact:
Email: support@phaethon.llc
Subject Line: SynapseIP Support Request
Website: https://phaethonorder.com