World

We will open the download next month (Sep 15th). The Python client is in PyPi.

  • Paramus World is a LLM-driven physical objects representation store, based on knowgledge graph (KG).
  • The system is able to learn and update itself from several information sources and text.
  • World autonomously model complex relationships and provide context to LLMs.
  • Paramus World observes information flow and takes conclusions what to add to the graph.

Example

1.) This text (Gmail) is connected:

2.) Then one-shot ‘Translated’ into:
1. Physical Entities (categorized by type):

   - People:
       - Dr. Thorsten Gressling
   - Organizations/Companies:
       - Gressling.com
   - Locations, addresses, places:
       - Lab 131
   - Products, devices, equipment:
       - Systag Reactor
   - Physical objects, materials, substances:
       - Aspirin (Substance)
       - Sulfuric Acid (Substance)

2. Relationships between entities:

   - Dr. Thorsten Gressling is part of 
     Gressling.com (relationship: partOf)
   - Dr. Thorsten Gressling uses 
     the Systag Reactor (relationship: uses)
   - The Systag Reactor is contained 
     in Lab 131 (relationship: contains)
   - The Aspirin synthesis involves 
     Sulfuric Acid (relationship: involves)
   - The Aspirin synthesis is conducted 
     in Lab 131 (relationship: conductedIn)

3. Key insights about the physical world mentioned in these emails:

   - Dr. Thorsten Gressling is conducting 
     an experiment involving the synthesis of Aspirin 
     catalyzed by Sulfuric Acid.
   - The experiment will be conducted using a Systag 
     Reactor located in Lab 131.

3.) And added to the RDF – database.
4.) Then part of the ‘World’

Chat: “How many reactors we have” – Answer: “There is one reactor in Lab 131”

The idea of a self-curating persistence layer for representing “what is there” – real-world entities like persons, positions, molecules, etc. – aligns conceptually with ontologies, knowledge graphs, and scene graphs. Obvious: LLM generate usefull answers if they know ‘what is out there’.

These serve as structured models of a persistent “world state” and are used in various domains including semantic web, robotics, LLM agents, and digital twins.

A server software to install

Pricing, Download and install

Free

Paramus World free tier provides a local knowledge‑graph representation of physical entities (people, substances, equipment) derived from text, enabling lightweight semantic context for small‑scale experimental projects.

free (1000 items)

SME

The SEM edition extends capacity to 50 000 entities and relationships, supporting mid-sized departmental or collaborative R&D projects with richer interlinking and inference density.

$5000 + 20%/yr (10k items)

Enterprise

Enterprise tier accommodates up to 1 million physical‑entity items, addressing industrial or multi‑institutional deployments with detailed world‑model representation.

Ask for details

Version 1 pricing

  • Free Edition: no cost — academic/academic‑access compatible.
  • SEM Edition: one‑time setup fee €5 000; maintenance 20% per year.
  • Enterprise Edition: Contact Thorsten Gressling

Why a worldview?

LLMs are semiotic systems: They operate in semiology, not ontology. They encode how humans talk about reality, not the reality itself.

The PARAMUS World application integrates an LLM with an RDF (Resource Description Framework) knowledge graph, allowing users to interact with structured data through natural language while maintaining a persistent semantic knowledge base.

Integrated JSON-RPC API

  • Secure Authentication: JWT Bearer tokens with rate limiting
  • JSON-RPC 2.0 Protocol: Full specification compliance
  • World Integration: API methods have access to world data
  • Available Methods:
    • chat.submit: Submit messages with world context
    • sparql.query: Execute SPARQL SELECT queries
    • sparql.update: Execute SPARQL UPDATE operations
    • system.health: Get server status and health
Scroll to Top