Vol. 01 — In Circulation
A Journal of AI Systems Architecture
Winter 2026 / ISSN 2749–0041

Designing systemsthat remember,recover & govern.

A model is a component, not a system. This volume documents the architecture around it — where reliability is actually decided.

Edited byHill Patel
ChaptersSeven
Reading≈ 11 min
§01ThesisField Notes

Great systems are not defined by model intelligence alone. They are defined by the architecture that surrounds it — the memory that persists, the pathways that recover, the orchestration that stays deterministic under load.

Intelligence is abundant and improving on a schedule no engineer controls. What remains scarce is structure: the discipline to make a capable model behave predictably, remember what it learned, and act within governed bounds.

Memoryisarchitecture.

Interlude — on persistence
§03Research QuestionsQ. 01–03
Q. 01

How should a system remember across sessions without accumulating noise, drift, or contradiction?

Persistence
Q. 02

When execution is interrupted, how does a system recover its exact position and continue as if uninterrupted?

Recovery
Q. 03

How do many agents, tools, and policies coordinate so the outcome stays deterministic and governable?

Orchestration
FIG.06 — PERSISTENT LATTICE
System 01 / Persistence

Memexa memory that keeps its shape.

A durable structure that lets a system carry what it learned across sessions, without drift.

Memory is stored as typed entities and relations, not a growing transcript. Writes reconcile against what exists; contradictions are resolved, not appended. A lattice, not a landfill.

10k+
Interactions before recall degrades
0drift
Reconciled writes, no silent contradiction
Typed entitiesReconciled writesTraversal recall
— INTERRUPTFIG.07 — CHECKPOINT & RESUME
System 02 / Recovery

Cairnpathways that survive interruption.

Deterministic checkpointing that resumes long-running work at the exact step it left, after any failure.

Every meaningful step leaves a marker — a cairn — with enough state to reconstruct the path. On crash or pause, recovery replays to the last marker and continues. No re-running from zero; no duplicated effects.

100%
Resumable at last committed step
<1s
Median time to re-attach and continue
CheckpointsIdempotent replayExactly-once
L3 · POLICYL2 · ROUTINGL1 · AGENTS
System 03 / Orchestration

Nexusdeterminism under coordination.

A layered orchestrator where agents, tools, and policy compose into one predictable, governable execution.

Coordination is arranged as nested layers: agents at the core, routing above, policy at the perimeter. Every action passes outward through governance before it reaches the world. Concurrency is real; nondeterminism is not.

4layers
Agents · routing · policy · audit
1:1
Same input, same execution plan
DeterministicPolicy-gatedAuditable

Systems rememberso peopledon't have to.

Interlude — on trust
§06The Plan — How It FitsFig. 09

One substrate,
three concerns.

Read left to right: a request enters, the orchestrator plans against persistent memory, executes through recoverable steps, passes every action through policy, and returns a result that can be replayed exactly.

REQUESTNEXUS · ORCHESTRATORdeterministic planMEMEX · MEMORYCAIRN · RECOVERYPOLICY · GOVERNANCEevery action auditedEXECUTE · TOOLSRESULT— REPLAY / RESUME
Fig. 09 — Reference architectureThe three systems are one plan seen from three angles — memory, recovery, and orchestration composing a single, replayable flow.

The model will keep getting better on its own. The parts that decide whether you can trust it — memory, recovery, governance — are the parts we still build by hand.

From the editor's note, Vol. 01