Introduction

akh-medu is a neuro-symbolic AI engine that combines hyperdimensional computing (Vector Symbolic Architecture) with knowledge graphs and symbolic reasoning. It runs entirely on the CPU with no LLM dependency, no GPU requirement, and no external NLP models.

What It Does

akh-medu stores, reasons about, and discovers knowledge. You feed it facts (triples like "Dog is-a Mammal"), and it can:

Infer new knowledge via spreading activation, backward chaining, and superposition reasoning
Reason symbolically using e-graph rewrite rules (equality saturation)
Search semantically using high-dimensional binary vectors (VSA)
Act autonomously via an OODA-loop agent with 15 built-in tools
Parse and generate natural language in 5 languages via a grammar framework
Serve knowledge over REST and WebSocket APIs

Architecture

┌──────────────────────────────────────────────────────────┐
│                      Engine API                          │
│  create_symbol · add_triple · infer · query · traverse   │
├──────────┬──────────┬──────────┬──────────┬──────────────┤
│   VSA    │Knowledge │Reasoning │Inference │   Agent      │
│  Ops     │  Graph   │  (egg)   │ Engine   │  OODA Loop   │
│ ─────────│──────────│──────────│──────────│──────────────│
│ HyperVec │petgraph  │ rewrite  │spreading │ 15 tools     │
│ SIMD     │oxigraph  │ rules    │backward  │ planning     │
│ ItemMem  │SPARQL    │ e-graphs │superpos. │ psyche       │
├──────────┴──────────┴──────────┴──────────┴──────────────┤
│                   Tiered Storage                         │
│  Hot (DashMap) · Warm (mmap) · Durable (redb)            │
└──────────────────────────────────────────────────────────┘

Key Differentiators

CPU-only: SIMD-accelerated hypervector operations (AVX2 where available, generic fallback everywhere else). No GPU, no CUDA, no model weights.
No LLM dependency: All reasoning is algebraic (VSA bind/unbind, e-graph rewriting, graph traversal). Grammar-based NLP replaces transformer models.
Hyperdimensional computing: 10,000-bit binary vectors encode symbols. Similarity is Hamming distance. Binding is XOR. Bundling is majority vote.
Full provenance: Every derived fact has a traceable derivation chain recording exactly how it was produced.
Multilingual: Grammar-based parsing and generation in English, Russian, Arabic, French, and Spanish with cross-language entity resolution.

How to Read This Book

Getting Started walks you through building from source and your first knowledge session.
Concepts explains the core data model and reasoning strategies.
Agent covers the autonomous OODA-loop agent.
Advanced dives into compartments, workspaces, grammars, and shared partitions.
Server documents the REST and WebSocket APIs.
Reference has the component status matrix and full CLI command reference.

Source Code

The source code is on GitHub: Toasterson/akh-medu

License

akh-medu is licensed under the GPLv3. For integration into proprietary applications, contact the author.

Keyboard shortcuts