How we work

A client journey, not three separate services.

We designed our services to build on each other. You don't need to buy all three pillars or all four courses at once. Most clients start with a health check. Startups selling AI to enterprise customers start with the Enterprise Readiness Check. Startups and SMEs implementing governance in their own business start with the AI Governance Health Check. Both grow from there.

01
Assess

Find out where you stand.

Health Check
or Enterprise Readiness Check
02
Prioritise

Identify the gaps and opportunities.

03
Govern

Build practical policies, roles and controls.

04
Enable

Build organisation-wide capability.

05
Improve

Assess where process improvement is worth pursuing.

06
Automate

Implement workflows that clear the bar.

07
Assure

Keep governance current as AI use changes.

STARTING POINT

AI Governance Health Check

For startups and SMEs using AI. We find out what AI is already in use, what the risks are, and what governance exists.

ENTERPRISE DEALS

Enterprise Readiness Check

The health check for startups selling AI to enterprises: the governance checks buyers expect, the gaps closed and the evidence documented.

ONGOING

Governance Advisory

As your business and AI use grow, we stay on to keep governance current.

The Edge AI Governance Cycle™

A practical process, not a certification or a compliance checklist.

Our own seven-stage method for identifying where AI is used, assessing what could go wrong, deciding what to do about it, putting safeguards in place, and keeping that judgement current as the business changes.

Stage 1: Identify 01 Identify Stage 2: Assess 02 Assess Stage 3: Decide 03 Decide Stage 4: Control 04 Control Stage 5: Operate 05 Operate Stage 6: Monitor 06 Monitor Stage 7: Improve 07 Improve EDGE AI PARTNERS The Edge AI Governance Cycle™ Govern. Enable. Automate.

  1. Identify
    What AI are we using?Know what AI exists, where it is used and who owns it.Typical activities
    • Build an inventory of AI tools, models, vendors and use cases
    • Find shadow AI through short staff surveys, software and expense reviews and team interviews
    • Record owner, users, data used and purpose for each entry
    • Set the governance baseline
    What it producesA comprehensive AI inventory, shadow-AI detection and a governance baseline.
  2. Assess
    What could go wrong?Understand the risk and impact of each use case before relying on it.Typical activities
    • Rate likelihood and impact for each use case
    • Review data protection, including whether personal data is involved
    • Review bias and fairness where outputs affect people
    • Review the vendor and third-party terms
    What it producesRisk exposure, impact analysis, bias and data privacy review.
  3. Decide
    Should we use it?Make a deliberate, recorded decision rather than letting use drift.Typical activities
    • Review the assessment with the accountable person or group
    • Decide: deploy, deploy with conditions, modify, or prohibit
    • Set conditions and a review date
    • Record who decided and why
    What it producesA formal, logged decision to deploy, modify or prohibit a use case.
  4. Control
    What safeguards are required?Put safeguards in place in proportion to the risk.Typical activities
    • Acceptable use and AI policies
    • Roles and accountability
    • Technical guardrails: access, configuration, data loss prevention
    • Human oversight points in each workflow
    What it producesPolicies, roles, technical guardrails and data protocols.
  5. Operate
    How do people use it?Make sure people use AI safely and consistently day to day.Typical activities
    • Train people at the right level (the four-course pathway)
    • Publish practical guidance and playbooks
    • Provide a support channel and a clear escalation route
    • Build human review into workflows where judgement is needed
    What it producesTrained, supported, safe day-to-day use.
  6. Monitor
    Is it still working safely?Confirm, on a set cadence, that AI is still performing safely.Typical activities
    • Track health metrics such as error rates, complaints and overrides
    • Run periodic audits and spot checks
    • Log and respond to incidents
    • Watch for vendor and model changes and regulatory developments
    What it producesHealth metrics, audits and incident logging.
  7. Improve
    What needs to change?Learn from what happened, update the system, and find processes that AI or process automation could improve. Then loop back to Identify.Typical activities
    • Review incidents, metrics and audit findings
    • Update controls, policies, training and systems
    • Retire or change AI that no longer earns its place
    • Feed new AI, changes and automation candidates back into Identify
    What it producesUpdated controls, policies and systems, then the Cycle loops back to Identify.
Two different models

Journey ≠ Governance Cycle™.

The client journey answers how an engagement with us unfolds.

The Edge AI Governance Cycle™ answers how your own business governs AI once governance is in place: Identify → Assess → Decide → Control → Operate → Monitor → Improve.

Ready to start?

Most clients start at step one.

Book a health check