AI governance that works.
An inventory of every AI system, risk tiers, policies and controls built into the tools your teams already use, so AI stays accountable without slowing delivery.
Free 30-minute scoping call. Written estimate within 5 working days.
Inventory · Risk tiers · Controls in your tools · Audit-ready
Illustrative interface. Values are sample data.
By the numbers
* Sample figures for layout review. Replace with audited numbers from engagement reports before launch.
Six services, one accountable AI programme.
AI governance is the set of roles, policies and controls that decide which AI you use, how it is checked, and who answers for it.
-
01
AI governance strategy.
Who decides what about AI, how decisions are recorded, and how governance fits your existing risk teams.
-
02
AI policy management.
Acceptable-use, procurement, development and incident policies, written for the people who must follow them.
-
03
AI risk assessment.
Every AI system inventoried and assessed for impact, with a risk tier and an owner.
-
04
Regulatory compliance.
A gap analysis against the EU AI Act, GDPR and sector rules, with a plan to close each gap.
-
05
Generative AI governance.
Rules for LLMs and copilots: approved tools, data handling, disclosure and output review.
-
06
Multi-model governance.
One set of controls across vendors, open-weight models and in-house models.
Four risk tiers, four levels of control.
Controls scale with risk. A spam filter shouldn't need the same paperwork as a hiring model, and tiers make that explicit.
-
Tier 1, Minimal: Record it and move on.
Spam filters, invoice extraction, internal search. Little impact on people.
- — Minimum controls —
- An owner and an entry in the inventory.
- Inventory entry
- Named owner
- Annual review
-
Tier 2, Limited: Tell people it's AI.
Chatbots, generated content and copilots that people interact with directly.
- — Minimum controls —
- People know they are dealing with AI.
- AI disclosure to users
- Output review rules
- Prompt and output logging
-
Tier 3, High: Test, document and oversee.
Hiring, credit, medical, safety and access to essential services.
- — Minimum controls —
- Documented risk management and human oversight before use.
- Risk and impact assessment
- Bias and accuracy testing
- Human oversight and appeals
-
Tier 4, Unacceptable: Don't build it.
Uses that are banned by law or against your values, such as social scoring.
- — Minimum controls —
- Blocked at intake, with the reason recorded.
- Stopped at intake
- Reason recorded
- Escalated if attempted
* Tiers follow the structure of the EU AI Act. Your counsel confirms how each system is classified.
Clear ownership from start to finish.
Governance fails when nobody owns the decision. Every activity gets one accountable owner, written down.
| Activity | Board | AI council | Product owner | Engineering | Legal and risk |
|---|---|---|---|---|---|
| Approve the AI strategy | A | R | C | I | C |
| Classify a new AI system | I | A | R | C | C |
| Run bias and accuracy tests | I | I | A | R | C |
| Approve a high-risk launch | I | A | R | C | C |
| Handle an AI incident | I | A | R | R | C |
| Review the risk register | I | A | C | I | R |
R responsible · A accountable · C consulted · I informed. A sample model; yours is agreed in the strategy work.
The standards we help you meet.
Most organisations answer to several at once. We map them onto one set of controls, so each requirement is met once.
— Regulations —
- EU AI Act
- GDPR
- CCPA
- HIPAA
— Frameworks —
- NIST AI RMF
- ISO/IEC 42001
- ISO/IEC 23894
- OECD AI Principles
— Sector and security —
- Model risk management (SR 11-7)
- PCI DSS
- SOC 2
- ISO/IEC 27001
* We help you align with these. Certification and legal sign-off stay with your auditors and counsel.
Accountable without the bureaucracy.
Good governance is mostly clarity: what AI may do, who checks it, and where the record lives.
-
01
What AI should and shouldn't do.
Clear rules on approved uses, banned uses and the data each system may touch.
-
02
Oversight that fits the work.
Review steps sized to risk and built into the tools teams use, not a separate portal.
-
03
Clear ownership.
Every AI system has a named owner, accountable for its performance and its risks.
-
04
A record of every decision.
Approvals, tests and incidents are logged, so you can show a regulator what happened and why.
-
05
Not legal advice.
We design and run the programme. Your counsel confirms the legal position on each system.
Illustrative interface. Values are sample data.
The first cohort is being written.
Named case studies publish here with the client's sign-off, real numbers and the honest limits. Want to be one of them?
Talk to us →Five documents you keep.
Written for the people who run the programme day to day, and for the regulator who may one day ask.
Seven weeks to a working programme.
Governance starts with what you already run, and ends inside the tools your teams already use.
-
01
Weeks 1–2
Find every AI system.
Interviews and tool scans build the inventory, including AI inside vendor software.
-
02
Weeks 2–3
Assess and tier.
Each system is assessed for impact and given a risk tier and an owner.
-
03
Weeks 3–5
Write the rules.
Policies, review steps and templates, written with the teams who will use them.
-
04
Weeks 5–6
Build it into your tools.
Intake forms, approvals and model cards in the ticketing and documentation tools you use.
-
05
Week 7
Test on real launches.
Two or three real AI launches go through the process, to find the friction.
-
06
Ongoing
Run and improve.
Quarterly reviews, incident handling and updates as regulations change.
AI governance for your industry.
Regulated industries carry the most risk, but every organisation using AI needs an owner and a record.
Mobility.
Dispatch, ETA and rider-support agents for ride-hailing and fleet platforms.
ETA models · Demand forecasting · Support agentsLogistics.
Route optimisation, document extraction and exception handling across the shipment lifecycle.
Route optimisation · Document AI · Exception triageFinance.
KYC document extraction, fraud signals and customer assistants that stay inside compliance rules.
KYC extraction · Fraud signals · Compliance copilotsReal estate.
Listing enrichment, valuation models and lead-qualifying agents for brokerages and proptech.
Valuation · Listing enrichment · Lead agentsRetail / Manufacturing.
Demand forecasting, visual inspection and inventory copilots on the shop and factory floor.
Forecasting · Visual inspection · Inventory copilotseCommerce / Consumer goods.
Product search, personalisation and post-purchase support agents that cite the order.
Semantic search · Personalisation · Support agentsTravel.
Itinerary assistants, dynamic pricing and disruption-handling agents for OTAs and operators.
Itinerary agents · Pricing · Disruption handlingLet's discuss your industry.
If the data is messy and the stakes are real, the method still applies. Bring the workflow; we'll bring the evaluation set.
Start a project →AI governance questions, answered.
Short answers here. Longer ones on the scoping call.
Message us on WhatsApp. A consultant replies within one working day.
What is AI governance?
AI governance is the set of roles, policies and controls that decide which AI systems an organisation uses, how they are checked, and who is accountable for them.
What's the difference between AI governance and data governance?
Data governance covers the data itself: quality, access and retention. AI governance covers what systems do with that data: purpose, testing, oversight and accountability. The two share controls, and we connect them.
How does AI governance support the EU AI Act and GDPR?
By inventorying your AI systems, classifying them by risk, and putting in place the documentation, testing and human oversight each tier requires. It supports compliance; your counsel confirms the legal position.
Does governance slow down AI delivery?
It shouldn't. Low-risk systems get a light touch, reviews are built into existing tools, and clear rules mean teams stop waiting for ad-hoc approvals.
Can governance cover chatbots and copilots?
Yes. Generative AI gets its own rules for approved tools, data handling, disclosure to users and review of outputs, scaled to the risk of each use.
How long does it take to set up AI governance?
A working programme usually takes six to eight weeks: inventory, risk tiers, policies and a process tested on real launches. Ongoing support is optional.
How much does AI governance consulting cost?
Most programmes start with a fixed-fee governance review, quoted on the scoping call. Further work is priced per phase, and every quote is written and fixed for its scope.
Do you provide legal advice?
No. We design and run the governance programme and map it to regulations and standards. Your legal counsel confirms how the law applies to each system.
Bring the workflow. We'll bring the plan.
Three steps from first message to kickoff. No deck, no commitment until you sign.
- 01Scoping call30 minutes with an engineer, not a salesperson.
- 02Written estimateAn evaluation plan, a timeline and a price within 5 working days.
- 03KickoffSign the scope and the discovery sprint starts.
Prefer to message directly? WhatsApp +91 7096010005
