People, apps & agents
Everyone and everything in the company that wants to use AI.
Why it mattersYou can't govern what you can't see. Most companies find far more AI in use than they expected.
02 / AI governance & controls
Run AI across your company, with every control in place.
Govern your data, write the rules into software and send every AI call through one gateway. Then AI can scale without surprises.
- Builds
- Data governance, policy layer, AI gateway
- Aligned to
- ISO 42001, EU AI Act, NIST RMF
- Ends with
- Controls running in production
01 / The architecture
What a governed AI company looks like.
01People, apps & agentsemployeesbusiness appsAI agents
02Policy layerwho may use whatapprovalsper-team rules
03AI gatewayroutingredactionspend limitslogs
04Models & agentsapproved modelsyour agentsswappable vendors
05Data governanceclassifiedownedaccess-controlledlineage
06 · Audit & evaluationlogs · evals · reports
Policy layer
Your AI rulebook, written into software instead of a PDF.
Why it mattersIt works like your spending-approval matrix, applied to AI. Rules are enforced automatically, not by memo.
AI gateway
One front door that every AI request passes through.
Why it mattersThink of a corporate card with limits and a monthly statement. You get control over cost, vendors and leaks in one place.
Models & agents
The AI itself: models from several vendors and the agents built on them.
Why it mattersYou can switch to a better or cheaper model without rewiring the business.
Data governance
The rules on what AI is allowed to know.
Why it mattersIt works like security clearance for information. Customer and financial data only reach the AI that is cleared for them.
Audit & evaluation
A flight recorder for every AI action, plus tests before anything ships.
Why it mattersWhen a regulator, auditor or board member asks what the AI did, you have the answer.
02 / What you get
Deliverables, not decks.
Data governance
Your data classified, owned and access-controlled, so AI only reads what it should.
Policy layer
Who may use which AI, on which data, with whose approval, enforced in software.
AI gateway
One door for every model call: routing, redaction, spend limits and logs.
Audit & evaluation
Every AI action recorded, tested before release, and reviewable later.
03 / Process
How it works
- Step 1DiscoveryLearn what your company needs and agree the AI plan.
- Step 2PlanningClassify data, roles, access, tools and security needs.
- Steps 3–4 · in parallelImplementationSet up governance, policy layer, evals, gateway, monitoring and access controls.TransformationMove teams and processes onto the new setup as each piece goes live.
- Step 5Verify & handoverTest every control, train your team and hand it all over.
Contact
Let’s talk.
Tell us what you want AI to do. You’ll hear back within a day.
Prefer to talk? Book 30 minutes directly.