NornicDB - V1 🎉

[orneryd]

NEW REPOSITOIRY STANDALONE

https://github.com/orneryd/NornicDB

🎉 NornicDB V1.0 Release Highlights

The Graph Database That Learns — Neo4j-compatible, GPU-accelerated, with memory that evolves.

---

🚀 What is NornicDB?

NornicDB is a high-performance graph database designed for AI agents and knowledge systems. It speaks Neo4j's language (Bolt protocol + Cypher) so you can switch with zero code changes, while adding intelligent features that traditional databases lack.

---

🐳 Get Started in 30 Seconds

# Apple Silicon (M1/M2/M3) with bge-m3 embedding model + heimdall
docker pull timothyswt/nornicdb-arm64-metal-bge-heimdall:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-arm64-metal-bge-heimdall


# Apple Silicon (M1/M2/M3) with bge-m3 embedding model
docker pull timothyswt/nornicdb-arm64-metal-bge:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-arm64-metal-bge

# Apple Silicon (M1/M2/M3) BYOM
docker pull timothyswt/nornicdb-arm64-metal:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-arm64-metal

# Apple Silicon (M1/M2/M3) BYOM + no UI 
docker pull timothyswt/nornicdb-arm64-metal-headless:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-arm64-headless

# NVIDIA GPU (Windows/Linux) with bge-m3 embedding model + heimdall
docker pull timothyswt/nornicdb-amd64-cuda-bge-heimdall:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-amd64-cuda-bge-heimdall

# NVIDIA GPU (Windows/Linux) with bge-m3 embedding model
docker pull timothyswt/nornicdb-amd64-cuda-bge:latest
docker run -d --gpus all -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-amd64-cuda-bge

# NVIDIA GPU (Windows/Linux) BYOM
docker pull timothyswt/nornicdb-amd64-cuda:latest
docker run -d --gpus all -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-amd64-cuda

# CPU Only (Windows/Linux) BYOM
docker pull timothyswt/nornicdb-amd64-cpu:latest
docker run -d -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-amd64-cpu

# CPU Only (Windows/Linux) BYOM + no UI
docker pull timothyswt/nornicdb-amd64-headless:latest
docker run -d --gpus all -p 7474:7474 -p 7687:7687 -v nornicdb-data:/data \
  timothyswt/nornicdb-amd64-cpu-headless

Open http://localhost:7474 — Admin UI with AI assistant ready to query your data.

✨ V1 Feature Highlights

🔌 Neo4j Drop-In Compatibility

• Bolt Protocol — Use any official Neo4j driver (Python, JavaScript, Go, Java, .NET)
• Cypher Query Language — Full support for MATCH, CREATE, MERGE, WITH, RETURN, etc.
• Schema Management — Constraints, indexes, and vector indexes
• Zero Code Changes — Your existing Neo4j applications just work

from neo4j import GraphDatabase
driver = GraphDatabase.driver("bolt://localhost:7687")
# That's it. Your Neo4j code works unchanged.

---

⚡ Performance (3-52x Faster Than Neo4j)

LDBC Social Network Benchmark (M3 Max, 64GB):

Query Type	NornicDB	Neo4j	Speedup
Message content lookup	6,389 ops/sec	518 ops/sec	12x
Recent messages (friends)	2,769 ops/sec	108 ops/sec	25x
Avg friends per city	4,713 ops/sec	91 ops/sec	52x
Tag co-occurrence	2,076 ops/sec	65 ops/sec	32x

Resource Efficiency:

• Memory: 100-500 MB vs 1-4 GB for Neo4j
• Cold Start: <1s vs 10-30s for Neo4j

---

🧠 Intelligent Memory System

Memory that behaves like human cognition with automatic decay:

Memory Tier	Half-Life	Use Case
Episodic	7 days	Chat context, sessions
Semantic	69 days	Facts, decisions
Procedural	693 days	Skills, patterns

MATCH (m:Memory) WHERE m.decayScore > 0.5
RETURN m.title ORDER BY m.decayScore DESC

---

🔗 Auto-TLP: Automatic Relationship Inference

NornicDB weaves connections automatically:

• Embedding Similarity — Related concepts link together
• Co-access Patterns — Frequently queried pairs connect
• Temporal Proximity — Same-session nodes associate
• Transitive Inference — A→B + B→C suggests A→C
• Edge Decay — Unused connections fade over time

# Enable automatic relationship inference
NORNICDB_AUTO_TLP_ENABLED=true

---

🎯 GPU-Accelerated Vector Search

Native semantic search with hardware acceleration:

• Apple Silicon — Metal Performance Shaders
• NVIDIA — CUDA acceleration
• Hybrid Search — RRF fusion of vector + BM25

CALL db.index.vector.queryNodes('memory_embeddings', 10, $queryVector)
YIELD node, score
RETURN node.content, score

OTHER GPUS TO BE SUPPORTED TOO AS SOON AS I FIND TIME

---

🤖 Heimdall AI Assistant

Built-in AI that understands your database:

• Natural Language Queries — "Show me system metrics", "Run health check"
• Bifrost Chat Interface — Real-time SSE streaming
• Plugin System — Custom actions with lifecycle hooks
• In-Memory llama.cpp — Direct SLM integration, no HTTP calls

NORNICDB_HEIMDALL_ENABLED=true ./nornicdb serve
# Access Bifrost at http://localhost:7474/bifrost

---

🧩 APOC Functions (60+ Built-In) (964 in the plugin)

Neo4j-compatible utility functions:

// Text processing
RETURN apoc.text.camelCase('hello world')     // "helloWorld"
RETURN apoc.text.slugify('Hello World!')      // "hello-world"

// Machine learning
RETURN apoc.ml.sigmoid(0)                     // 0.5
RETURN apoc.ml.cosineSimilarity([1,0], [0,1]) // 0.0

// Collections
RETURN apoc.coll.sum([1, 2, 3, 4, 5])         // 15

Plus plugin system — drop .so files for custom extensions.

---

🛡️ Production-Ready Features

• WAL (Write-Ahead Log) — Durability with crash recovery
• Snapshots — Binary gob encoding for fast persistence
• Audit Logging — Track all operations
• RBAC — Role-based access control
• Edge Provenance — Track why edges were created
• Evidence Buffering — Aggregate signals before materializing

---

📦 Docker Images Ready to Go

Platform	Image	Description
Apple Silicon	nornicdb-arm64-metal-bge-heimdall	Full - Embeddings + AI
Apple Silicon	nornicdb-arm64-metal-bge	Standard - With BGE-M3
Apple Silicon	nornicdb-arm64-metal	Minimal - BYOM
NVIDIA GPU	nornicdb-amd64-cuda-bge-heimdall	Full - CUDA + AI
NVIDIA GPU	nornicdb-amd64-cuda-bge	Standard - With BGE-M3
CPU Only	nornicdb-amd64-cpu	No GPU required

30-Second Start:

docker run -d -p 7474:7474 -p 7687:7687 \
  timothyswt/nornicdb-arm64-metal-bge-heimdall:latest

---

🔧 Feature Flags (Runtime Configuration)

All experimental features are opt-in:

Feature	Default	Flag
Auto-TLP	❌ Off	NORNICDB_AUTO_TLP_ENABLED
Edge Decay	✅ On	NORNICDB_EDGE_DECAY_ENABLED (requires NORNICDB_AUTO_TLP_ENABLED=true)
GPU Clustering	❌ Off	NORNICDB_GPU_CLUSTERING_ENABLED
Heimdall QC	❌ Off	NORNICDB_AUTO_TLP_LLM_QC_ENABLED

---

🌐 MCP Integration

Model Context Protocol tools for AI agents:

• store — Create nodes with embeddings
• recall — Retrieve by ID or filter
• discover — Semantic search
• link — Create relationships
• task / tasks — Task management
• index / unindex — Codebase indexing

---

📊 V1 by the Numbers

• 3-52x faster than Neo4j across benchmarks
• 60+ APOC functions built-in
• 6 Docker images for different platforms
• 30+ feature flags for fine-tuning
• <1s cold start vs 10-30s for Neo4j
• 100-500MB RAM vs 1-4GB for Neo4j

---

🚀 Get Started

# Apple Silicon with everything
docker run -d -p 7474:7474 -p 7687:7687 \
  -v nornicdb-data:/data \
  timothyswt/nornicdb-arm64-metal-bge-heimdall:latest

# Open http://localhost:7474

---

📚 Documentation

• Quick Start Guide
• Feature Reference
• User Guides
• API Reference

---

🙏 Feedback Welcome!

We'd love to hear from you:

• What features would you like to see?
• What's working well?
• What could be improved?

---

"The Norns weave your data's destiny" — NornicDB automatically discovers and creates the connections that give your data meaning.

[orneryd]

now I gotta work on moving it to another repo but I'll get to that after i fix other features in mimir and after people actually test it and get a chance to play with it and give me real feedback from the wild.

[orneryd]

new repo https://github.com/orneryd/NornicDB

[straga]

@orneryd
If I switch from Neo4j to NordicDB for Mimir, will it work from that box?

[orneryd]

it should! i am in the process of hardening and finishing up nornic itself to where i want it and then i will be back on enhancing Mimir