Skip to content

regen-network/koi-research

Repository files navigation

KOI Research Repository

Overview

This repository contains the Knowledge Organization Infrastructure (KOI) implementation for RegenAI, integrating BlockScience's KOI v3 protocol with Regen Network's metabolic ontology to create a distributed knowledge management system.

🚀 Current Status (September 2025)

100% Complete - Full KOI sensor-to-agent pipeline operational and tested

✅ Implementation Complete - BREAKTHROUGH ACHIEVEMENT

  • Complete KOI Pipeline: End-to-end flow from sensors to agents fully operational
  • KOI Event Bridge v2: Real-time processing with RID-based deduplication and versioning
  • BGE Server Integration: 1024-dimensional embeddings generated via HTTP API
  • PostgreSQL Direct Storage: Immediate integration with Eliza agent database using isolated tables
  • Real-time Processing: Content processed and available within seconds of ingestion
  • Agent RAG Access: Processed content immediately accessible for agent queries
  • CAT Receipt Generation: Complete transformation provenance tracking operational
  • Production Deployment: Full pipeline tested and verified with real content processing
  • Performance Metrics: Real-time processing, immediate agent availability, complete audit trails

✅ Research Foundation

  • Apache Jena Fuseki integration with 3,851+ RDF triples
  • D3.js interactive graph visualization (326+ entities)
  • Metabolic entity extraction from 1,100+ documents
  • JSON-LD to TTL conversion pipeline
  • Ontological research and unified taxonomy development

📖 Master Documentation

🎯 KOI Master Implementation Guide - PRIMARY REFERENCE

Complete consolidated guide (Version 1.1) covering:

  • KOI v3 protocol implementation with FULL/PARTIAL nodes
  • Sensor, processor, and coordinator node architecture
  • Entity resolution and deduplication strategies
  • Complete end-to-end knowledge flow
  • Integration with ElizaOS agents
  • Next phase: KOI network implementation

📚 Supporting Documents

Note: Individual documents preserved for reference, but Master Guide is the primary source of truth.

Key Recommendations

  1. Dual Identification System: RIDs for semantic identity + CIDs for deduplication
  2. Transformation Provenance: Track every operation with CAT receipts
  3. Apache Jena RDF Store: Semantic reasoning with W3C standards compliance
  4. Separation of Concerns: Standalone KOI Processor separate from agents
  5. Phased Implementation: MVP in 2 weeks, full system in 12 weeks

Implementation Timeline

  • Weeks 1-2: Foundation (RID/CID, CAT receipts)
  • Weeks 3-4: Processing Pipeline (modular stages, cost optimization)
  • Weeks 5-6: Knowledge Graph (Apache Jena + RDF deployment)
  • Weeks 7-8: Sensor Network (real-time monitoring)
  • Weeks 9-10: Agent Integration (all 5 agents on KOI)
  • Weeks 11-12: Registry Framework Integration (SPARQL endpoints, federation)

Repository Structure

koi-research/
├── README.md                    # This file
├── KOI_COMPLETE_RESEARCH.md    # Main comprehensive document
└── sources/                     # Research sources
    └── blockscience/
        ├── koi/                 # KOI main repo
        ├── koi-net/            # KOI-net implementation
        └── rid-lib/            # RID library

Quick Start

To begin implementation:

  1. Review the complete research document
  2. Set up KOI Processor Node (Week 1)
  3. Implement dual identification system
  4. Create transformation receipts
  5. Test with one agent

Related Projects

Current Status

🚀 BREAKTHROUGH ACHIEVEMENT - Complete KOI sensor-to-agent pipeline operational ✅ Production Deployment - KOI Event Bridge v2 with deduplication processing sensors through BGE to PostgreSQL ✅ Real-time Processing - Content flows from sensors to agent access in seconds ✅ BGE Integration - 1024-dimensional embeddings via dedicated HTTP API server ✅ Agent Integration - Direct RAG query access to processed KOI content ✅ Complete Provenance - CAT receipts track every transformation step ✅ Research Foundation - Ontological framework supporting production implementation 🎯 100% Complete KOI Pipeline - Production Ready and Tested

Key Achievements - Production-Ready KOI Pipeline

  • Complete KOI Implementation: Full sensor-to-agent pipeline operational and tested
  • KOI Event Bridge v2: Real-time processing with RID-based deduplication and versioning
  • BGE Server Integration: Dedicated HTTP API generating 1024-dimensional embeddings
  • PostgreSQL Direct Storage: Immediate agent database integration with pgvector
  • Real-time Agent Access: Content available for RAG queries within seconds of processing
  • CAT Receipt System: Complete transformation provenance tracking operational
  • Production Architecture: FastAPI-based services with async processing and error handling
  • Research Foundation: 36-class unified ontology with OWL compliance
  • Knowledge Graph: Apache Jena integration with 3,851+ RDF triples
  • End-to-End Testing: Complete pipeline verified with real content processing

Technical Stack

  • Apache Jena + Fuseki SPARQL server for RDF storage and reasoning
  • OWL ontologies for semantic modeling and automated inference
  • SPARQL endpoints for Registry Framework integration
  • JSON-LD for semantic web compatibility
  • Provenance tracking with RID/CID dual identification

For Regen Network's regenerative AI infrastructure initiative

About

KOI Research - Knowledge Organization Infrastructure research and development

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 14