Skip to content

Agentic AI governance · Assessments

Bring the hard question to the table.

A human-led assessment of agent permissions, tool use, and oversight. Bring the sources into view, clarify accountability, and agree practical next actions.

Request a walkthrough Explore the approach

Human-led approach · Sources, responsibilities and next actions

Three colleagues discussing a governance review, facing one another across a shared table.
Review dossierAgentic AI GovernancePermissions · Tools · Oversight

Frame the governance question: what can the agent decide, access, or do?

Illustrative scene · Sample review material

PAID AI USE-CASE READINESS ASSESSMENT

Bring one use case. Prepare the next decision.

A paid, human-led review of one AI workflow: what supports the next step, what must close first and who is accountable. Receive a documented recommendation and action plan, with an optional platform handover.

Explore the paid assessment →Read the guide and working records ↗

The preparation materials are free. Assessment scope, timing, fee and any platform subscription are agreed separately.

The engagement

Start with the decision that needs clarity.

A focused human-led assessment helps a leadership, risk, or operating team make sense of an AI workflow, its supporting material, and the responsibilities around it.

Questions to bring to the first conversation

Are responsibilities clear?

Who owns the system, reviews its output, accepts the risk, and follows up when something changes?

What do the sources actually support?

Which system, vendor, and control records are available? Where are the gaps, outdated versions, or assumptions?

What needs a decision next?

Which observations need leadership attention, an operating change, or further specialist review?

From a question to a reviewable record.

These stages are shaped around your agreed scope. Follow the sample example, or select a stage to inspect it.

Choose the decision, workflow, or AI system to examine. Agree who needs to participate and which references belong in scope.

Illustrative review record1 / 4

Question

Where does human review enter the decision?

A sample team uses an AI recommendation to support a business decision. The proposed review asks who checks the recommendation, when they intervene, and what is recorded.

In scope
One workflow and its review responsibilities
To agree
Decision owners, review questions and reference points
QuestionSourcesObservationNext action

Sample material shows the method. It is not a client finding or a delivered assessment.

Work products with a purpose.

These public working previews illustrate possible outputs. Engagement deliverables are agreed in a written proposal.

Systems → Owners → Context

AI registry and risk observations

A starting inventory of systems, their purposes, accountable people, and recorded risk considerations.

Open preview

Observation → Sources → Review

Evidence and control gap pack

An observation linked to its sources, the limits of the review, and the question that remains open.

Open preview

Portfolio → Priorities → Decisions

Executive AI risk brief

A leadership view of the scoped portfolio and the decisions requiring attention, with a supporting appendix.

Open preview

Action → Owner → Follow-up

Action register

Proposed work, accountable owners, and the evidence needed to review closure. Use the private Customer Portal to explore the sample action records.

Open preview

People carry the context.
People own the decisions.

A useful review depends on the client team and reviewer working through the same question with clear responsibilities.

Your team

  • Identify the business question and responsible stakeholders.
  • Agree which material can be shared and explain its context.
  • Review observations, set priorities, and retain decisions and risk acceptance.

The reviewer role

  • Frame the scope and references with your team.
  • Organize sources, document observations, and make limitations visible.
  • Bring proposed actions back for review and agree the handover.

The first conversation

Bring the question.
Agree the work.

Start with a walkthrough of the public example and a discussion of what your team needs to understand.

Request a walkthrough

Before work begins

A written proposal confirms scope, stakeholders, source handling, timing, fees, and deliverables. Continuing support is agreed as part of the proposal.

Assessment boundaries

The assessment documents observations against agreed references. It does not provide a certification, legal conclusion, or assurance opinion. Specialist advice may be needed for decisions outside the agreed scope.

After the initial review

Your team reviews proposed actions and agree a handover or further work. Use the private Customer Portal to explore sample records in your workspace and prepare the next conversation.

Use the first conversation to describe your needs. Do not send confidential records through the public site.