Make public-sector knowledge easier to understand and easier to act on.

Entitlement, statute and procedure, applied consistently in every office and explained plainly to the citizens affected by them.

Caseworkers and citizens read one body of rules

Public bodies hold knowledge written to be legally precise, and it has to serve caseworkers and citizens alike. Caseworkers need to apply it identically across offices. Citizens need to understand what they are entitled to, what is required of them and what happens next, without a background in administrative law.

The rules themselves are layered: primary legislation, secondary legislation, departmental policy, operational guidance and local practice. Each layer has its own owner, its own change cycle and its own interpretation, and frontline staff sit at the bottom of the stack resolving the differences in real time.

Add the structural pressures. Services span multiple departments and agencies, staff turnover is real, contact volumes are seasonal and unforgiving, and every decision may be appealed, reviewed or reported on.

Frontline staff carry the reconciliation. They take a citizen’s circumstances, work out which layer of rules governs them, and produce a decision or an explanation, usually within a target handling time and always in a form that could be reviewed afterwards.

What a public body keeps in its Brain:

  • Entitlement rules and their conditions
  • Operational guidance, tied to the legislation it implements
  • Decision guidance, with the reasons a decision records
  • Evidence requirements for each application
  • Statutory deadlines and the clocks they start
  • Fees and charges for each service
  • Appeal and review procedures
  • Rules in plain language, approved for citizens
  • Accessibility requirements for each service
  • Handover rules between departments and agencies
  • Delegated authority: who may decide what
  • Local practice where it legitimately differs

What it costs when guidance is unclear

Inconsistent decisions on comparable facts are the central risk. They generate appeals, reviews and complaints, they consume the time of experienced staff, and they damage public confidence in a way that takes years to rebuild.

The service costs are immediate: repeat contacts when the first answer did not fit, failed digital journeys, avoidable handovers between offices, and backlogs that grow whenever a rule changes.

There is a fairness dimension that outweighs the operational one. People with the same circumstances should receive the same outcome, and when they do not, the cause is unclear guidance applied conscientiously in different ways.

  • The challenge:

    Variation between offices

    The same circumstances producing different outcomes in different locations.

    With ClearMash Brain:

    Consistent decisions

    Comparable cases treated alike, wherever they are handled.

  • The challenge:

    Repeat contact

    Citizens coming back because the first answer did not fit their case.

    With ClearMash Brain:

    Fewer repeat contacts

    Citizens get an answer that fits their circumstances the first time.

  • The challenge:

    A gap between rule and practice

    New guidance reaching frontline practice long after it took effect.

    With ClearMash Brain:

    Faster implementation of change

    New guidance reaches frontline practice in days.

Public-sector AI needs governed knowledge

Public bodies are introducing assistants into citizen contact and internal support, in an environment where an incorrect statement about entitlement is a public matter.

That requires knowledge with a named owner, an effective date, a plain-language expression that has been approved, a defined boundary where a case must reach a human decision-maker, and a record of what the service said, and when, because someone will ask.

Where accessibility obligations apply, an approved plain-language expression of a rule serves the citizen, the caseworker and the assistant equally well, and it exists because somebody owns it.

What a caseworker has to settle before deciding

  • WHATWhat this citizen is entitled to, given their circumstances.
  • WHICHWhich guidance version applied on the date of the application.
  • WHOWho is responsible, and what transfers at the handover.
  • WHENWhen the statutory clock starts and stops.
  • WHYWhy the decision was reached, in language the citizen can be given.
  • HOWHow every office reaches the same outcome on the same facts.
A citizen at the counter, and the layers of rules that could apply. The date of the application pins the version that governs it, and the decision leaves in both forms - the reason and the steps for the file, and the plain words the citizen in front of you is given.

The same facts, and a decision that may be reviewed years later.

The reason is recorded, and the citizen is told in plain words.

A citizen at the counter, and the layers of rules that could apply. The date of the application pins the version that governs it, and the decision leaves in both forms - the reason and the steps for the file, and the plain words the citizen in front of you is given.

What Entitlement · Statute · Guidance

  • Guidance note
  • Guidance v12
  • Entitlement
  • Statutory rule
  • Policy layer
  • Local practice
  • Evidence list
  • Threshold
  • Appeal route
  • Plain wording
  • Time limit

How Casework · Reasons · Handover

  • Decision steps
  • Evidence check
  • Reason recorded
  • Who decides it
  • Handover point
  • Clock starts
  • Refusal wording
  • Escalation rule
  • Review path
  • AI agent boundary
  • Notify steps

Context Decides which What and which How to act on, for this case

  • Who This applicant
  • When Applied on
  • Where This office
  • Which Version then
  • Why Reason to give
  • Caseworkersdecide the same way
  • Citizenstold it in plain words
  • AI in channelsinside the guidance
  • One wording, every counter

Tangible impact

The changes a working Brain makes, in the order a public counter meets them.

Business impact

  • Reduction in inquiry handling time 28%
  • Reduction in compliance violations 14%
  • Reduction in new employee onboarding 70%
  • Increase in employee satisfaction 62%
  • Increase in self-service adoption 30%
  • Increase in customer satisfaction 15%
  • Increase in first contact resolution (FCR) 40%
  • Reduction in expert escalations 65%

Technological impact

  • AI agent accuracy from 85% to 97%
  • Reduction in AI costs 74%
  • Increase in AI agent speed 25%

Customer result

At a large municipal authority, call times fell 50%.

Additional measures

And the measures a public body already publishes about its own service:

  • appeal and review rates
  • repeat contact
  • time to implement guidance
  • consistency between offices

Accountability by design

Named ownership, approval before publication, version history, access control and a record of what was in force at any date. A citizen’s personal data stays in departmental systems of record. Certifications and security documentation are published on the Security page. Where a service is delivered jointly by several bodies, ownership of each subject can be assigned explicitly, so every rule has an owner.

See one rule reach the caseworker’s decision and the citizen’s answer

Book a demo and we will show you an entitlement rule with its effective date, and the recorded reason and the plain words that both come from it.

Prefer to start with the numbers? The impact estimator works them out