AIngleSemantic infrastructure
AIngleAIngle

The world's first semantic DAG .

A directed acyclic graph with meaning that fuses a distributed ledger, graph database, and edge computing into a single platform. Every link is signed — a verifiable history for intelligent applications.

1.ª semantic DAG in the world·<1MB Edge node (IoT)·3 APIs · REST·GraphQL·SPARQL
View on GitHubExplore capabilities
firmado ✓ ed25519[Subject] → [Predicate] → [Object]

A graph with meaning, and with a signature

◇Why AIngle

Not a database, not a blockchain. Something new.

Legacy systems can't keep up

IoT, real-time compliance, and intelligent automation outpace traditional databases.

Databases are slow; blockchains are heavy

One doesn't understand context; the other doesn't scale. And custom solutions create silos.

Data, logic, and privacy, siloed

Each layer in its own system. Integrating and auditing becomes costly and fragile.

◇Solutions

From knowledge to compliance, at the edge.

Deep Context

Preserve institutional knowledge: architecture decisions, the why behind the code, and their semantic relationships — searchable forever.

Semantic Compliance

Real-time AML/KYC: monitor sanctions lists, detect hidden networks with graph analysis, and leave an immutable trail.

Edge Intelligence

IoT without compromise: sub-second decisions on <1MB devices, with P2P gossip and embedded intelligence.

◇Capabilities

A unified platform, not a stack of parts.

Semantic graph engine
Native graph database with SPARQL queries. Model complex relationships and traverse connections without external dependencies.
Zero-knowledge privacy
Prove facts without revealing data: Schnorr, Pedersen commitments, and Bulletproofs built in.
Kaneru — multi-agent
Multi-agent execution system with reinforcement learning (Q-Learning, SARSA, TD) for autonomous decisions.
Unified API
One interface, three protocols: REST, GraphQL, and SPARQL. The Cortex API adapts to your needs.
Smart contracts
Rust DSL compiled to WASM. Typed, isolated, deterministic execution: deploy logic that runs anywhere.
Real-time visualization
Interactive D3.js panel: watch your DAG evolve, filter, search, and analyze the system visually.
◇Grounded retrieval

Context with citation, provenance, and proof.

AIngle turns your documents into cited context anchored to provenance. Every retrieved fragment is tied to the signed DAG action that recorded it — auditable end to end. And every answer carries a groundedness:

grounded

Grounded

Strong similarity and sufficient corroborating sources — the answer is well supported.

weak

Weak

There is relevant material, but below the corroboration threshold — answer with caution.

ungrounded

Ungrounded

No source supports it — the sources cannot sustain an answer.

Exposed via MCP: aingle_ingest · aingle_ground · aingle_sources

◇Production-ready

High availability and consensus.

Multi-node cluster with Raft consensus: replicas on every node, fault-tolerant durability, and TLS encryption between nodes.

  • Raft consensus
  • Write-Ahead Log
  • Streaming snapshots
  • Inter-node TLS
  • Quorum reads
◇Integrate it

Official SDKs + MCP.

Connect AIngle from your favorite language, or plug it into Claude via Model-Context-Protocol.

  • JavaScript / TypeScript
  • Python
  • Go
  • Swift
  • Kotlin
  • MCP
Dual: Apache-2.0 + Commercial

Build on verifiable infrastructure.

Free for personal use, education, and research. For commercial use we evaluate each case — tell us yours and we'll find the right fit.

View on GitHubCommercial license