Skip to content

IndentiaPlatform/IndentiaDB

BranchesTags

Repository files navigation

IndentiaDB

Multi-Model Database for Knowledge Graphs & Enterprise AI

License: Apache-2.0 Rust Version: 0.1.0 SPARQL 1.2 RDF 1.2

IndentiaDB unifies RDF graphs, JSON documents, full-text search, vector embeddings, and labeled property graphs into a single ACID-compliant engine. Built in Rust. Native SPARQL 1.2, Elasticsearch-compatible API on port 9200, first-class AI/RAG integration — all in one binary.

Feature Overview

Feature Description
Multi-Model Five data models: RDF + JSON + LPG + Full-Text + Vector — no polyglot persistence needed
SPARQL 1.2 W3C Working Draft compliance, RDF-star (Triple Terms), quoted triples, SEMIJOIN/ANTIJOIN
SurrealQL SQL-like document + relational queries with graph edge traversal
Enterprise Search BM25 + Dense vector hybrid search with Bayesian fusion (NDCG@10: 0.9149)
Elasticsearch API Drop-in compatible on port 9200 — works with existing ES clients, Kibana, dashboards
AI-Ready / RAG Knowledge graph extraction, HNSW vector indexing, RAG pipelines, chunking
Real-Time LIVE SELECT via WebSocket, DEFINE EVENT triggers, alerting engine
ACID Transactions Snapshot isolation across all 5 data models simultaneously
Security Three-layer ACL: LDAP/OIDC/JWT auth + RBAC/ABAC + triple-level RDF-star ACL
Kubernetes Native Operator with CRD, StatefulSet management, auto-scaling, rolling upgrades
High Availability Raft consensus (OpenRaft), leader election, automatic failover, 3-node cluster
SHACL Validation W3C SHACL Core + SPARQL constraints, shapes graphs, REST validation endpoint
GeoSPARQL 60-bit geospatial encoding, nearest neighbor, distance joins, WKT parsing
Semantic Inference RDFS/OWL forward-chaining materialization with incremental updates
Bitemporal Time-travel queries with transaction time + valid time (AS OF, BETWEEN)
Federated Queries SPARQL SERVICE clause, R2RML virtual graphs, SQL database federation
WASM Support Browser & edge runtime via WebAssembly
QLever Compatibility FSST decompression, 6-permutation index (SPO/SOP/PSO/POS/OSP/OPS)

Quick Start

Docker (simplest)

docker run -p 7001:7001 -p 9200:9200 ghcr.io/indentiaplatform/indentiadb-trial:latest

IndentiaDB is now running with:

  • SPARQL / GraphQL / REST / WebSocket on port 7001
  • Elasticsearch-compatible API on port 9200

Docker Compose

git clone https://github.com/IndentiaPlatform/IndentiaDB.git
cd IndentiaDB
docker compose up -d

Binary

# Download the latest release for your platform
curl -L https://github.com/IndentiaPlatform/IndentiaDB/releases/latest/download/indentiagraph-$(uname -s)-$(uname -m) \
  -o indentiagraph
chmod +x indentiagraph

# Start the server
./indentiagraph serve --port 7001

Verify

# Health check
curl http://localhost:7001/health

# Run a SPARQL query
curl -X POST http://localhost:7001/sparql \
  -H "Content-Type: application/sparql-query" \
  -d "SELECT * WHERE { ?s ?p ?o } LIMIT 10"

# Use the Elasticsearch-compatible API
curl http://localhost:9200/_search \
  -H "Content-Type: application/json" \
  -d '{"query": {"match": {"content": "knowledge graph"}}}'

Architecture

┌──────────────────────────────────────────────────────────────┐
│                          Clients                              │
│   SPARQL · SurrealQL · GraphQL · REST · ES API · WebSocket   │
└───────────────────────────┬──────────────────────────────────┘
                            │
┌───────────────────────────▼──────────────────────────────────┐
│                    IndentiaDB Server                          │
│                                                               │
│  ┌─────────────┐ ┌─────────────┐ ┌───────────────────────┐  │
│  │  SPARQL 1.2 │ │  SurrealQL  │ │  ES-Compatible API    │  │
│  │   Engine    │ │   Engine    │ │  (port 9200)          │  │
│  └──────┬──────┘ └──────┬──────┘ └───────────┬───────────┘  │
│         │               │                     │               │
│  ┌──────▼───────────────▼─────────────────────▼───────────┐  │
│  │              Hybrid Query Router                         │  │
│  │     SPARQL() + LPG() + SurrealQL unified routing        │  │
│  └─────────────────────────┬───────────────────────────────┘  │
│                            │                                   │
│  ┌─────────────────────────▼───────────────────────────────┐  │
│  │                  Multi-Model Store                        │  │
│  │                                                           │  │
│  │  RDF Triples  │ JSON Documents │ LPG Graph Edges         │  │
│  │  Vectors      │ Full-Text Index│ Geospatial              │  │
│  └─────────────────────────┬───────────────────────────────┘  │
│                            │                                   │
│  ┌─────────────────────────▼───────────────────────────────┐  │
│  │         Storage Backend (SurrealDB / TiKV)               │  │
│  │    6-permutation RDF index · HNSW vector · BM25          │  │
│  └──────────────────────────────────────────────────────────┘  │
└────────────────────────────────────────────────────────────────┘

Data Models

IndentiaDB natively supports six data models through a unified storage engine:

Model Query Language Example
Relational SurrealQL SELECT name, salary FROM employee WHERE department = "Engineering"
Document SurrealQL CREATE project CONTENT { name: "Alpha", tags: ["ai", "graph"] }
Graph (RDF) SPARQL 1.2 SELECT ?name WHERE { ?p a foaf:Person ; foaf:name ?name }
Graph (LPG) LPG JSON DSL LPG({"kind":"traverse","start":{"label":"Person"},"edge":"knows"})
Vector SurrealQL SELECT *, vector::similarity::cosine(embedding, $query) AS score FROM chunks
Full-Text ES Query DSL {"query": {"match": {"content": "knowledge graph"}}}

Comparison

Feature Comparison

Feature IndentiaDB QLever (C++) Oxigraph Blazegraph Apache Jena MarkLogic ArangoDB PostgreSQL SQL Server Oracle
Language Rust C++ Rust Java Java C++ C++ C C C
Clustering ✅ Raft HA ❌ ❌ ✅ ❌ ✅ ✅ ✅ ✅ ✅
Multi-Modal ✅ 5 models ❌ RDF only ❌ RDF only ❌ RDF only ❌ RDF only ✅ ✅ ⚠️ ext ⚠️ ext ⚠️ ext
SPARQL 1.1 ✅ ✅ ✅ ✅ ✅ ✅ ❌ ❌ ❌ ❌
SPARQL 1.2 ✅ Full ✅ Partial ⚠️ Partial ❌ ❌ ❌ ❌ ❌ ❌ ❌
RDF 1.2 ✅ Full ⚠️ Partial ⚠️ Partial ❌ ❌ ❌ ❌ ❌ ❌ ❌
RDF-star ✅ Native ⚠️ Partial ✅ ❌ ✅ ❌ ❌ ❌ ❌ ❌
GeoSPARQL ✅ ❌ ❌ ⚠️ ✅ ext ✅ ✅ ✅ ext ✅ ✅
Full-Text Search ✅ BM25 native ✅ ql:contains ❌ ✅ Lucene ✅ Lucene ✅ ✅ ⚠️ tsvector ✅ ✅
Federated Queries ✅ SERVICE ✅ SERVICE ✅ SERVICE ✅ SERVICE ✅ SERVICE ⚠️ ❌ ❌ ❌ ❌
Vector Search ✅ HNSW ❌ ❌ ❌ ❌ ❌ ✅ ✅ ext ❌ ❌
Elasticsearch API ✅ Port 9200 ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌
RAG / AI ✅ Native ❌ ❌ ❌ ❌ ❌ ⚠️ ❌ ❌ ❌
Bitemporal ✅ ❌ ❌ ❌ ❌ ✅ ❌ ❌ ❌ ❌
Reasoning ✅ RDFS/OWL ❌ ❌ ✅ ✅ ✅ ❌ ❌ ❌ ❌
Real-Time Events ✅ LIVE/Events ❌ ❌ ❌ ❌ ❌ ✅ ❌ ❌ ❌
K8s Operator ✅ ❌ ❌ ❌ ❌ ✅ ✅ ✅ ✅ ✅
WASM ✅ ❌ ✅ ❌ ❌ ❌ ❌ ❌ ❌ ❌
License Apache 2.0 Apache 2.0 MIT/Apache GPL 2.0 Apache 2.0 Commercial Apache 2.0/Commercial PostgreSQL Commercial Commercial

Security Comparison

Security Feature IndentiaDB QLever Oxigraph Blazegraph Apache Jena MarkLogic ArangoDB PostgreSQL SQL Server Oracle
LDAP ✅ ❌ ❌ ❌ ❌ ✅ ✅ ✅ ext ✅ ✅
OIDC / OAuth2 ✅ ❌ ❌ ❌ ❌ ✅ ✅ ❌ ✅ ✅
JWT Bearer ✅ ❌ ❌ ❌ ❌ ✅ ✅ ❌ ✅ ❌
HTTP Basic Auth ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅ ✅
RBAC ✅ ❌ ❌ ✅ ❌ ✅ ✅ ✅ ✅ ✅
Graph-Level ACL ✅ ❌ ❌ ✅ ❌ ✅ ✅ ✅ ✅ ✅
Triple-Level ACL ✅ RDF-star ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌
Windows SID Support ✅ ❌ ❌ ❌ ❌ ❌ ❌ ❌ ✅ ✅
Rate Limiting ✅ ❌ ❌ ❌ ❌ ✅ ✅ ❌ ✅ ✅
Audit Logging ✅ ❌ ❌ ❌ ❌ ✅ ✅ ✅ ext ✅ ✅
FROM Clause Filtering ✅ auto ⚠️ ⚠️ ⚠️ ⚠️ ✅ ❌ ❌ ❌ ❌
RDF-star Security ✅ ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌ ❌

Performance

SPARQL Benchmark (Olympics Dataset, January 2026, Apple M2 Max)

System Average Query Time vs Oxigraph
QLever (C++) 52ms 34x faster
IndentiaDB (Rust) 344ms 5x faster
Oxigraph (Rust) 1,764ms baseline

IndentiaDB achieves 5x faster SPARQL query execution than Oxigraph through its 6-permutation RDF index (SPO/SOP/PSO/POS/OSP/OPS) combined with FSST compression, while adding multi-model capabilities that pure SPARQL engines do not offer.

Hybrid Search Benchmark (SQuAD, NDCG@10)

Scorer Algorithm NDCG@10
RRF Reciprocal Rank Fusion (BM25 + Dense) 0.8874
Bayesian Bayesian Probability Calibration (BM25 + Dense) 0.9149

The Bayesian scorer auto-calibrates from the BM25 score distribution, requiring zero manual tuning. Enable it with ES_HYBRID_SCORER=bayesian.

Reference: "Bayesian BM25: A Probabilistic Framework for Hybrid Text and Vector Search"

API Endpoints

Endpoint Port Method Path Description
SPARQL Query 7001 GET/POST /sparql Execute SPARQL 1.2 queries
SPARQL Update 7001 POST /update SPARQL UPDATE operations
Graph Store 7001 GET/PUT/POST/DELETE /data SPARQL Graph Store Protocol
GraphQL 7001 POST /graphql GraphQL queries and mutations
SHACL Validation 7001 POST /shacl/validate Validate RDF against SHACL shapes
Health 7001 GET /health Server health + cluster status
Metrics 7001 GET /metrics Prometheus-compatible metrics
WebSocket 7001 WS /ws LIVE SELECT streaming
Alerts 7001 GET/POST /alerts Alerting engine REST API
ES Search 9200 GET/POST /_search Elasticsearch-compatible search
ES Index 9200 PUT /{index} Create/configure index
ES Document 9200 POST /{index}/_doc Index a document
ES Bulk 9200 POST /_bulk Bulk index/update/delete
ES KNN 9200 POST /{index}/_search Vector/KNN search

Configuration

Minimal Configuration (TOML)

[server]
bind_address = "0.0.0.0"
port = 7001

[surrealdb]
endpoints = ["ws://localhost:8000"]
namespace = "indentia"
database = "main"
username = "root"
password = "root"

[elasticsearch_compat]
enabled = true
host = "0.0.0.0"
port = 9200
hybrid_scorer = "bayesian"        # rrf | bayesian | linear
vector_search_mode = "hnsw"       # hnsw | brute_force

[cluster]
enabled = false
node_id = "node-1"
# seed_nodes = ["node-2:7002", "node-3:7003"]

Key Environment Variables

Variable Default Description
INDENTIADB_CONFIG config.toml Path to TOML configuration file
INDENTIADB_PORT 7001 Main server port
INDENTIADB_BIND_ADDRESS 0.0.0.0 Bind address
SURREAL_ENDPOINT — SurrealDB WebSocket endpoint
SURREAL_USERNAME root SurrealDB username
SURREAL_PASSWORD root SurrealDB password
ES_ENABLED true Enable Elasticsearch-compatible API
ES_PORT 9200 Elasticsearch API port
ES_HYBRID_SCORER rrf Hybrid search scorer (rrf, bayesian, linear)
ES_VECTOR_SEARCH_MODE hnsw Vector index type (hnsw, brute_force)
CLUSTER_ENABLED false Enable Raft cluster mode
CLUSTER_NODE_ID — Unique node identifier
CLUSTER_SEED_NODES — Comma-separated seed node addresses
QUERY_SPARQL_12 true Enable SPARQL 1.2 features
QUERY_INFERENCE false Enable RDFS/OWL inference
LOG_LEVEL info Log level (trace, debug, info, warn, error)

Deployment

Environment Guide
Docker / Docker Compose docs/deployment/docker.md
Kubernetes docs/deployment/kubernetes.md
OpenShift / OKD docs/deployment/openshift.md

Examples

Category Count Description Path
SurrealQL 25 CRUD, aggregates, transactions, events examples/surrealql/
RDF / SPARQL 8 Triples, named graphs, inference, federation examples/rdf/
LPG Graph 10 Traversals, PageRank, shortest path examples/lpg/
Enterprise Search & RAG 25 BM25, vectors, RAG, AI agents examples/enterprise-search/
Live Data & Agents 25 DEFINE EVENT, reactive patterns, agents examples/live-data/
Multi-Model Combined 5 All 5 models in hybrid queries examples/combined/

Documentation

Full documentation is available at dbdocs.indentia.ai.

Section Link
Getting Started dbdocs.indentia.ai/getting-started/
Configuration Reference dbdocs.indentia.ai/getting-started/configuration/
SPARQL 1.2 Guide dbdocs.indentia.ai/query-languages/sparql/
SurrealQL Guide dbdocs.indentia.ai/query-languages/surrealql/
Enterprise Search & RAG dbdocs.indentia.ai/features/enterprise-search/
Security & ACL dbdocs.indentia.ai/security/
API Reference dbdocs.indentia.ai/api-reference/
Deployment dbdocs.indentia.ai/deployment/

Contributing

Contributions are welcome. Please read CONTRIBUTING.md before opening a pull request.

License

Licensed under the Apache License, Version 2.0.

Copyright 2024-2026 Indentia B.V.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

Ownership and licence status

Indentia-authored original code and additions are owned by Indentia B.V., while use is governed by the applicable LICENSE files. See NOTICE or INDENTIA_NOTICE.md and OWNERSHIP.md.

About

Multi-model database combining RDF/SPARQL, SurrealQL, and Elasticsearch in a single query layer — built for Knowledge Graphs and Enterprise AI. → https://dbdocs.indentia.ai

Topics

rust elasticsearch database sparql rdf knowledge-graph graph-database full-text-search multi-model bm25 hnsw lpg surrealdb enterprise-ai bb25

Resources

Readme

License

Apache-2.0 license

Contributing

Contributing
Activity
Custom properties

Stars

6 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages