ARCHITECTURE OVERVIEW — AIONdb: Architecture of a Shared Memory Substrate →
AIONdb Explained →

The continuity engine for coordinated intelligence.

Agents are built to reason, but they lack the infrastructure to finish the work correctly — verified, coordinated, and provable. AIONdb gives autonomous agents the shared memory and exact ground truth they need to execute tasks, checking every command before it reaches a real system, and leaving a provable record.

The foundation for autonomous action is finally here.

Apply for Beta Speak with the Founder

WHAT IT IS

Memory your agents can act on.

AIONdb holds one shared ground truth: a live record of exact facts, kept in the order they happened and governed by who may see and change them. Agents come and go, models get swapped, and sessions end, but the record persists: queryable, provable, and audited.

The memory doesn’t live in the agent; the agent lives in the memory.

AIONdb doesn’t replace the systems you have. It connects to them and holds one live, exact, provable picture your agents act from.

THE DIFFERENCE

Watching, alerting, and predicting don’t get the work done.

Databases capture what happened, dashboards raise alerts, and AI tools make predictions. But all of them stop where the work begins: none of them will carry a task through to finished. The work waits for whoever notices it, and when no one does, it simply doesn’t happen.

AIONdb is built for that moment. It gives autonomous agents a shared, persistent memory and standing rules that live in the substrate itself — watching the record and firing the moment a condition is met — so a task is tracked until it’s finished, escalated if it’s ignored, and closed with signed proof. Done, not sent. Proven, not promised.

And orchestration frameworks wire agents together through a coordinator that becomes the single point of failure. AIONdb has no coordinator: agents coordinate through the shared record itself — one goes down, the rest keep working, and it catches up from the log when it returns. Zero lines of coordination code.

The engine enforces follow-through across three pillars:

EXACT

Facts, not the closest match.

Ask the same question, get the same answer: deterministic to the integer, citable to the source.

PROVABLE

A record you can verify.

Every write is signed, ordered, and chained into a tamper-evident record. A third party can verify the record without trusting us.

GATED

Checked before it acts.

Before any command reaches a real system, whether it’s deploying code or adjusting a clinical device, a deterministic check confirms the target is present and ready, or the command does not fire.

THE EVIDENCE

Every figure has a receipt.

Every figure below is a conservative floor, reproducible from timestamped records. The reproduction commands ship with every evaluation install.

270M+
Queries executed
Across 1,777+ timestamped reports
440B+
Error-free results returned
(result triples)
0
Failures
Across the record
799K+
Results / second
Single-machine record — set on a laptop
328/328
W3C SPARQL 1.1
Conformance suite
149
Patent-pending claims
Three USPTO provisional filings

Full production security path (authentication, TLS, role-based access, encrypted audit) at under 3% throughput cost, effectively zero at scale.

THE MECHANISM

Write. Verify. Gate.

Every action an agent takes travels the same three-step path.

Blocked · target not ready
passed  ·  blocked at the gate — the command never fires
  1. 01 · WRITE → delivers EXACT

    Agents record observations and decisions as typed facts, through the same authenticated path as every other client — and the policy engine bounds every participant to exactly what it may see and what it may share. No privileged shortcut, no roaming the record.

  2. 02 · VERIFY → delivers PROVABLE

    The engine orders every change and keeps the record coherent before anyone reads it. Every write is signed into the tamper-evident chain.

  3. 03 · GATE → delivers GATED

    When an agent’s command would touch a real system, a protocol-level probe confirms the target is live and ready before it passes.

Every task ends the same way: with signed proof it happened. Any moment can be replayed, so you can see exactly what occurred and why.

WHO IT’S FOR

Built for the teams putting agents to work.

Engineers reviewing an agent system across multiple laptops

Engineers building agent systems

Accepted teams are running the same afternoon: a single 30 MB binary, no dependencies, and what you test locally is what runs in production.

Clinical monitoring systems beside a patient in a hospital ward

Teams shipping consequential workflows

The memory, coordination, and audit that let a pilot survive production.

An operations engineer with a live energy-grid dashboard in a control room

Engineered for the highest-stakes environments

Hospital wards, buildings, energy grids — where a wrong action has consequences you can’t undo: the reason the gate exists.


Built on open standards. No data lock-in, no model lock-in.

One graph, reachable over every open surface your stack already speaks. Your data is portable, verifiable, and yours; swap AI models freely, and the memory stays in the substrate — a single engine, built in Rust.

REST GraphQL gRPC MCP W3C RDF SPARQL 1.1 OIDC

WHITE PAPER
The Shared Substrate
How intelligence has always coordinated — and why AI needs to learn the same lesson.
By Mike Monteith · March 2026

From cave paintings to enterprise databases to social networks — every advance in human coordination is the same story: someone built a better shared surface for knowledge. AI agents need the next one. This paper explains what it looks like, why it matters, and how AIONdb built it.

Read our white paper →
The Shared Substrate
Read online

About AIONdb

The AI world is asking urgent questions about agents — how they coordinate, how they remember, and what happens when they act. AIONdb was built to answer them.

AIONdb is a reactive knowledge graph and continuity engine — one shared ground truth that AI agents and people work from together. It was engineered from the memory layer up, with no existing database underneath: lock-free, zero-copy, and cryptographically auditable. It carries 149 patent-pending claims across three USPTO provisional filings, and it has been tested through 270 million+ queries producing 440 billion+ error-free result triples, with zero failures. Agents built on AIONdb keep their memory across sessions, model swaps, vendors, and restarts — and every command that would touch a real system is checked before it fires.


The Founder

AIONdb was created by Mike Monteith, a Canadian technology builder, inventor, and systems architect. For 35 years, Mike has been focused on how distributed, autonomous systems coordinate truth — eighteen of them chasing this exact problem. Mike founded and led ThoughtWire, a Canadian company focused on digital twin technology, for over 15 years, building real-time operating platforms for life-critical infrastructure at scale — smart hospitals, large commercial environments, and complex physical systems.

AIONdb is the culmination of that career. The gate exists because Mike spent those years in places where a wrong action can’t be undone.

Mike Monteith — Founder & Chief Architect

Want to learn more?

Is there a workflow you don’t trust an agent with today?

This is the one to bring to us. Apply for the beta, or write to the team.

Apply for Beta info@aiondb.ai