Aran AI helps government agencies and regulated organisations use AI safely. Aran Shield governs every AI request you make. Aran Cloud gives you a sovereign place to build and run your own AI agents.
Aran (அரண்) is Tamil for “fortress”.
The problem
Every agency is working out AI on its own: its own rules, its own logs, its own approvals. That means paying for the same work over and over, and gaps nobody can see.
Staff are already using AI tools, often without anyone knowing. Leaders are asked to prove AI is being used safely, but have nothing to show.
Aran gives you one control point, so you can see your AI use, govern it, and prove it.
Use Aran Shield on its own to govern the AI you already use. Add Aran Cloud when you want to build and run your own agents, with Shield built in.
A sovereign AI governance gateway. Every AI request passes through one control point that applies your rules, protects sensitive data, asks a person when it matters, and keeps a record you can prove.
A sovereign platform for building and running AI agents. Your teams build agents with the skills they already have. We take care of the containers, deployment, networking and security, with Aran Shield deciding what can safely leave New Zealand.
Aran Shield
Shield sits between your organisation and the AI models it uses. It works the same way whichever models you choose, so you stay in control instead of the big providers.
See it → Legitimise it → Prove it
Five parts
Every AI request goes through one door. Your rules and sensitive-data checks run before anything leaves, and each request goes to the right model at the right cost.
See who is using AI, what for, and what it costs, including tools staff already use without approval.
A tamper-evident record of every AI decision. Each entry is chained to the last, so any change shows.
A person approves high-risk actions before they happen. AI does the work; people make the calls that matter.
Every AI agent is tied to a named person who is responsible for it, with a kill switch to stop it at once.
Your cloud or ours. Run Shield inside your own cloud tenancy, or let us host it for you. The licence is the same either way.
Aran Cloud
In development
Getting even one AI agent into production on generic cloud services can take months. Aran Cloud gives New Zealand organisations a ready platform to build, deploy and run agents, with the security and data controls already in place.
Build agents and sub-agents in Python, starting there. No special platform skills needed.
Containers, deployment, networking and security are handled for you.
Local models handle everything they can. Work goes to offshore models only when Aran Shield says it is safe.
Use it as a service or in your own tenant: on Azure NZ North, NZ-owned clouds, or your own hardware.
Principles
Safe AI should not be only for the biggest organisations.
Every claim is backed by a record you can check.
Your data, your jurisdiction, your rules.
You stay in charge of your agents, not the big providers.
One place to set and enforce AI policy, instead of thirty.
See the cost of AI use, request by request.
Māori data sovereignty
Our role is to make sure the rules are followed, every time.
Contact
See Aran Shield working on a real scenario, or talk to us about Aran Cloud for your organisation.