Technology

Keep the history separate from the model.

Cairn is designed so the system's trusted history, sources and recovery process do not live only inside the AI model that happens to be running today.

SYSTEM ONLINE
EVENT HISTORY · APPEND-ONLY
SOURCES · LINKED
CHECKPOINTS · VERIFIABLE
MODEL CHANGE · SUPPORTED
How it fits together / 01

The model can change. The continuity layer stays with the system.

The AI model does the thinking. Cairn keeps the history and recovery record underneath it.

CAIRN / REFERENCE ARCHITECTURE● CONTINUITY LAYER ONLINE
AI model layer
MODEL ACurrent model / provider / host
→
TRUSTED HISTORY + RECOVERYCAIRN CONTINUITYKept separate from any one model
→
MODEL BReplacement / upgrade
What Cairn keeps
01

Preserved event history

Important events stay in order. New information is added instead of silently replacing the old record.

02

Source tracing

Important claims can point back to where they came from, who or what created them, and when.

03

Original records preserved

Original files and artifacts remain independently available instead of being replaced by summaries.

04

Rebuild current state

The system can rebuild what is true now from verified history instead of trusting a mutable summary.

05

Verified checkpoints

Recovery points can be checked before the system resumes after a move, failure or restore.

06

Model change & recovery

A replacement AI model can resume from the same trusted history without having to copy the old model word for word.

What that enables
History cannot be silently overwritten

The original record stays intact.

Integrity can be checked

Important objects and checkpoints can be verified for unexpected changes.

Recovery can move

Trusted state can be carried across hosts and providers.

Failure is part of the design

The system is built with restoration and long-term change in mind.

Core pieces / 02

Keep the original record first. Build summaries and working memory on top of it.

The point is simple: a later summary should never become the only version of what happened.

01

Append-only event history

Important events are added in order with links to what came before, where the information came from and integrity data that helps detect tampering.

02

Original files stay addressable

Original documents and artifacts remain available so a later summary does not replace the source material.

03

Memory linked back to evidence

Working memory, summaries and conclusions can point back to the records that support them.

04

Verified checkpoints

Recovery points can be checked before the system resumes after a move, failure or restore.

05

State reconstruction

The current working state is rebuilt from verified history instead of being trusted just because a summary says it is correct.

06

Model transition

A replacement model can recover prior state and unfinished commitments without having to imitate the old model word for word.

What Cairn is not / 03

This is more than adding memory to a chatbot.

Memory tools are useful. Cairn is focused on what happens when the model changes, the system moves or recovery has to be trusted.

A

Not a personality prompt

Telling a new model who it is does not prove that it recovered the old system correctly.

B

Not just vector memory

Search and retrieval help with recall, but they do not automatically preserve the order of events or which record should be trusted.

C

Not just workflow checkpointing

Restarting a process is different from carrying years of history across a model replacement.

CAIRN CONTINUUM

See what parts of this architecture have already been tested.

Research