Operational memory in practice: how teams stop useful knowledge being lost

The organisation may have learned the lesson once, but that does not mean the lesson is easy to reuse the next time. That is where operational memory becomes useful.

Operational memory is not just about storing documents somewhere an AI system can read them. It is about helping teams reuse practical knowledge in a more consistent way. It turns lessons, repeated checks and hard-won judgement into something the wider organisation can work from.

Why useful knowledge gets lost

Most organisations do not lose expertise because it disappears overnight. They lose it because it stays informal. It sits in people’s heads. It lives in email trails, handover notes, old reports, marked-up documents and conversations that made sense at the time but were never turned into something reusable. The business still has the information somewhere, but not in a form that helps people act on it when they need it.

That is why the problem is not just storage. It is retrieval and reuse. A record can exist without becoming useful memory. A lesson can be written down without becoming part of how the wider team works.

A simple example

Imagine a head of safety who has spent years reviewing contractor RAMS across several sites.

Over time, they build strong judgement. They know that one contractor often writes polished documents with vague responsibilities. They know that another tends to copy emergency arrangements from previous jobs. They know that certain types of maintenance work regularly create the same weak points around isolation, access control or handover.

None of this is mysterious. It is simply repeated experience. Now imagine that this person is on leave, moves roles, or leaves the organisation altogether.

The documents are still there. The occurrence logs are still there. Previous reviews may even be saved somewhere. But the practical judgement about what usually goes wrong, what needs checking first, and which issues matter most is suddenly much harder for the wider team to use. That is the gap operational memory is meant to close.

It helps turn repeated judgement into something that can be retrieved and reused by more than one person.

What operational memory looks like in practice

In practical terms, operational memory may include things such as:

• recurring weak points in contractor submissions
• examples of what good looks like
• site-specific lessons from previous incidents
• repeated follow-up questions that reviewers tend to ask
• patterns that only become obvious across multiple entries or reviews
• decisions that were taken last time and why they were taken

The important point is that this is not just archive material. It is working knowledge. When someone asks a question, reviews a document or faces a similar situation again, the organisation should be in a better position to retrieve what it already knows.

Why this matters

This matters because many operational teams are under pressure to do more with less. That usually means there is less time for slow knowledge transfer, less time for one experienced person to review everything themselves, and less tolerance for repeating the same mistakes because useful lessons were not easy to find. Operational memory helps in three simple ways.

1) Reduces dependence on a few individuals. The organisation becomes less fragile if more people can work from the same lessons and practical checks.

2) Improves consistency. A stronger site or team should not be the only place where good judgement exists.

3) Helps newer staff work from a better base. They may not have years of experience yet, but they can still be supported by what the organisation has already learned.

What operational memory is not

It is not a magic memory for everything or a replacement for judgement. And it is not just uploading documents into an AI system and hoping for the best.

The value comes when useful knowledge is shaped around real work. That means making it easier to retrieve what matters, in a form people can actually use, at the point where the task is being done. That may be during a review, a handover, an incident, a planning discussion or a briefing.

A better way to think about it

A simple way to think about operational memory is this:

• A record tells you what happened.
• Operational memory helps you make better use of what the organisation has already learned.

That is a much more practical asset. It means the business is not just storing experience. It is getting better at reusing it.

Why this is becoming more important

As AI becomes more capable, more organisations will be tempted to think of memory as a storage problem. That is too narrow.

The more useful question is whether the organisation can retrieve the right lesson, the right pattern, the right previous check or the right practical example when the task is live. That is why operational memory matters. The point is not simply to remember more. It is to make hard-won judgement easier to reuse in real work.

Stop useful knowledge being lost

If useful judgement still depends too heavily on a few experienced people, operational memory may be the next step worth taking seriously.

SIRV AI helps teams retrieve procedures, evaluate documents, triage incoming information and retain lessons learned with an operational layer around AI use.

Frequently asked questions

What is System 3 thinking?
In this article, it means AI acting as a third layer in decision-making by helping retrieve, structure and present information for people. That usage is drawn from recent work by Steven Shaw and Gideon Nave.

What is cognitive surrender?
It is Shaw and Nave’s term for handing over too much judgement, effort or responsibility to AI when the system appears capable and convincing.

Why does this matter in high-consequence work?
Because overreliance on automation has long been associated in human factors research with weaker monitoring and reduced error detection, and similar concerns now apply to AI-supported decisions.

What is the role of an operational layer?
It is the practical structure around AI use that helps keep support bounded, checkable and useful in real work rather than allowing unchecked delegation. 

Author bio: Andrew Tollinton

Andrew Tollinton is CEO and Co-Founder of SIRV, which builds operational AI for safety, security and resilience teams. He focuses on practical, controlled AI use in serious environments, with particular interest in evidence, accountability and human judgement. Andrew chairs the Institute of Strategic Risk Management’s AI in Risk Management Special Interest Group and speaks regularly on AI governance and operational resilience.