Octimine

Summary
Provider

Dennemeyer Octimine

Fee
Fee paying
ASPI
No
Database description

Octimine is a software platform for searching, reviewing, and monitoring patents. It combines multiple search methods, including proprietary machine learning models, with AI-powered review tools based on large language models, and built-in collaboration features. Designed for teams across innovation, it helps users with different levels of IP expertise turn patent data into practical insight.

Highlights:

• Flexible patent search: Combine natural language, semantic search using machine learning, Boolean logic, and citation-based exploration to adapt to different use cases and levels of complexity
• AI-assisted document review: Summarize, compare, and extract key insights using Octi, your assistant powered by LLMs
• Collaboration and workflow tools: Work in shared spaces with tagging, comments, ratings, and project-specific visibility settings
• Custom monitoring profiles: Track technologies, competitors, or legal status changes and receive automated alerts
• Integration-ready: Use Octimine’s API to embed search and data access into your internal tools and workflows

Highlights
Legal status
Search results
Similarity search
Coverage
Patent Collections
  •  
  •  
  •  
Interfaces
Language
English
General search tools
General search tools
Cross-lingual semantic search
Search history queries
Semantic search
Similarity search

Octimine provides multiple search methods that can be used individually or in combination: semantic, Boolean, citation, and non-patent literature (NPL).

Semantic Search:

Octimine’s semantic search uses machine learning to match queries by meaning. Queries in any language are automatically translated into English in the background and searched against the global database. Features include boosted search (prioritizing keywords or CPC/IPC codes) and smart query expansion (AI-suggested reformulations of short queries for improved results).

Boolean Search:

Supports traditional Boolean operators, enhanced with NLP functions. Users can also sort results by similarity to a reference patent and combine Boolean queries with semantic and filtering options.

Citation Search:

Enables exploration of forward and backward patent citations, with tools for identifying prior art, mapping technological relationships, and analyzing influence within a field.

Non-Patent Literature (NPL) Search:

Allows searching NPL using English text queries or patent numbers to identify related scientific and technical publications.

Biological and chemical search tools
Biological and chemical search tools

Biological and chemical searches are possible through semantic search.

Classification
Classifications
Cooperative Patent Classification
International Patent Classification
Other

Schmölz patent technological classification

Value-added data
Value added data
Harmonized titles and/or abstracts
Standardized applicant names

Owner & Owner Subsidiaries information available

Patent family data
Patent family data
Artificial
Simple

Patents are grouped by DOCDB family number, with artificial families created where bibliographic data is incomplete.

Index lists
Index lists
Applicant name
Corporate tree
Invention text
Results
Results type
Annotation
Flagging
Keyword in context (KWIC) highlighting
Machine translation
Sorting by relevance
Translated titles and/or abstracts

Octi, Octimine’s AI assistant, is integrated into the results list for interactive review. Users can engage with Octi through a chat to request explanations, comparisons, or clarifications.
Each result includes an AI-generated summary of the patent and its relevance to the search query.
A claims match function highlights the specific passages in the claims that correspond to the user’s query.

Analysis
Analysis data
Applicant and/or inventor
Citation
Classification
Dates
Invention text
Analysis
Graphical
Statistical

Interactive charts
Custom visual dashboards
AI assistant Octi to review results or a single patent

Alerts
Alerts
Legal status
Search results

Octimine provides monitoring functions to track new publications and legal status changes, with flexible alerting and distribution options.

New Publications Monitor:

Monitoring profiles can be created in three ways:
• Semantic profiles: defined with patent numbers or text descriptions, useful for monitoring similar technologies.
• Boolean profiles: based on Boolean criteria such as keywords, companies, applicants, or classifications.
• Deep learning profiles: trained with large patent portfolios (tens or hundreds of documents) to detect new publications similar to the provided set.
These profiles continuously scan new patent data, and users can choose the method that best fits their monitoring objective. Results can be stored in folders, shared with colleagues, and delivered by email at a preferred frequency.

Legal Status Monitor:

Legal status profiles track changes to patents or families based on entered publication or application numbers. Numbers can be added manually, imported from a portfolio, or uploaded as a file. Each profile can cover up to 10,000 families. Users choose whether to track only the entered patents or all family members globally. Notifications of status changes, such as grants or expirations, are delivered in the platform and by email at the chosen frequency.

Data export
Data Export
Bibliographic data
Citations
Classification
Description and/or claims
Title and/or abstract

Exports are possible as PDF, XLSX or CSV files

Collaboration tools
Collaboration tools
Shared workfiles

Octimine provides built-in collaboration functions to support team-based patent search and monitoring.

Users can:
Save, tag, export, comment, highlight, and share documents directly in the platform.
Input from colleagues can be requested (e.g., ratings on specific results), and results can be archived in shared project spaces for structured review.
For scaled collaboration, the platform provides visibility and hierarchy settings to manage access and roles across teams.

Non-patent data
Non-patent content
Non-patent literature
Operators
General operators
Boolean (AND, OR, NOT, XOR)
Comparison (e.g. >, =<)
Proximity (within range)
Special

Text fields
• CONTAINS / DOES NOT CONTAIN – include or exclude documents with the specified term(s).
• AND / OR – logical operators; must be written in capital letters. Parentheses () control precedence.
• Implicit AND – multiple words without operators must all be present.
• “…” – exact phrase search.
• “… ”~N – proximity search; allows up to N words between phrase terms.
• ^N – boost operator; increases weighting of a term in ranking.

Dates
• = (EQUAL) – matches the specified date.
• <> (NOT EQUAL) – excludes the specified date.
• < (BEFORE) – earlier than the specified date.
• > (AFTER) – later than the specified date.

Numbers (e.g., publication, application)
• = (EQUAL) – matches exactly the entered number.
• <> (NOT EQUAL) – excludes the entered number.
• STARTS WITH / CONTAINS – partial matching of structured numbers.#

Entities (e.g., applicants, owners, inventors)
• EQUALS – exact name match.
• STARTS WITH – matches names beginning with the specified string.
• CONTAINS / DOES NOT CONTAIN – include or exclude documents with the specified string in the entity name.
• (Automatic name standardization is applied to unify common variants.)

Classifications (e.g., CPC, IPC)
• CONTAINS – retrieves documents with the specified classification code, including all child codes.
• DOES NOT CONTAIN – excludes documents with the specified code.

Authority / Jurisdictions
• EQUALS – restricts to the specified authority or jurisdiction.
• DOES NOT CONTAIN – excludes a specified authority.

Other fields (e.g., tags, legal status, custom attributes)
• Field-dependent operators (typically EQUALS / CONTAINS / DOES NOT CONTAIN) apply according to the structure of the metadata.

Truncation
Center truncation
Left truncation
Right truncation


Bibliographic data – At least one of the following is available: publication number, filing number, grant number, publication date, filing date, grant date, applicant name, assignee name, inventor name, patent classification.
Partial text – At least one of the following (but not all) are available: title, abstract, description, claims.
Full text – All of the following are available (where applicable for the given jurisdiction): description, claims