Digital Twin

Context that stays when the person is busy, away, or gone.

A digital twin is a context-aware representation of a person, a role, a team or the whole organisation. It works from approved knowledge, inside boundaries somebody set deliberately, and it acts only through workflows that already exist.

Start free trialTalk to us about a twin
What a twin is made of

Four scopes, one shape.

01
Person
The way one person works: their context, their judgement calls, the tasks they would rather delegate.
02
Role
What the job requires regardless of who holds it — so a handover is not a rebuild.
03
Team
How a group actually coordinates: who is consulted, what gets escalated, where work waits.
04
Organisation
The shape of the company as it really operates, rather than as the org chart draws it.
The problem it solves

Every AI conversation in your company starts from zero.

A model has no idea who you are, what your team decided last quarter, or which customer must never be emailed on a Friday. So people re-explain their world every single time — and when someone leaves, everything they knew about how the work is done leaves with them.

Approved knowledge only

A twin draws on sources somebody explicitly allowed. It cannot go looking for more.

Bounded action

What it may do is defined per twin: which workflows, which systems, which limits.

Learns from real work

Context comes from the work happening in the Workspace, not from a questionnaire.

Answers in your absence

Colleagues get the context they would have asked you for, without waiting for you.

Handover that holds

A role twin carries the working knowledge of the job to whoever takes it next.

Visible to its subject

The person a twin represents can read their own record. Nothing is built behind their back.

Typical use cases

Where it earns its place.

Delegating the work you dislike

The routine half of a job, handed to something that already knows the context it needs.

Covering absence properly

Holiday and handover stop meaning that a whole area of knowledge goes offline.

Answering "what if this person leaves"

A real answer, based on the work, instead of a guess and a hope.

Seeing the organisation as it works

Where decisions actually wait, and which dependencies nobody had named.

The principle

A mirror the organisation holds up to itself — not a periscope pointed at people.

Who can see what, whether a person can read their own record, and what cannot be extracted even with administrator rights — these are decided in how the capability is built, not in how it is sold. A twin that employees cannot trust is a twin nobody will give context to, and a twin without context is worthless.

The subject can read it
Whoever a twin represents can see what it holds about them.
Aggregate for the organisation
Organisation-level insight comes from aggregate views, not from reading individuals.
Not a performance tool
Twins are not designed to score, rank or monitor people, and will not be sold as that.
Boundaries survive admins
What cannot be extracted stays inaccessible regardless of privilege level.
Security and control

Control is what makes a twin acceptable at all.

Explicit knowledge scope
Sources are allow-listed per twin. Nothing is included by default.
Action boundaries
A twin acts only through approved workflows, never directly against a system.
Gateway permissions
Model access and data rules are inherited and cannot be widened locally.
Subject access
The person or team represented can review their own record.
Extraction limits
Individual-level records are not exportable, including by administrators.
Auditability
Every action a twin takes is attributed and retained like any other identity.
How security worksISO 27001 and ISO 9001 certified. Certificates and scope statements on request.
How it connects

Part of one platform.

Add a twin where the context matters most.

Start with one role or one team, on the work where losing context costs the most. Fourteen days, no card.

14 days · no payment card · bring your own provider account