LuminorixYou stay in control. Here is exactly how.
AI agents draft the next step. A person on your team approves or sends. Every action is on the record, and where an action can be reversed you can undo it. We also tell you what we do not have yet.
Send 2 reminders. Hold 1 for you.
You read each draft first. Where an action can be reversed, you can undo it.
Example card on example data. Nothing here is connected to a real system.
What an agent may do alone, and what it may not.
Read, draft, prepare
Look things up, write a draft, line up the next step.
Send, post, pay, change a price
A card with the reason. Approve or hold, and undo where it can be reversed.
Move money, write to someone new, delete
This is our rule for agents. A spoken yes never approves; you tap.
A change is checked before it reaches you.
Changes to an agent go through a separate reviewer step where it matters, before they go live. A failed check is reported as it is, never as a pass. Each release has a backup and a rollback ready.
- Built in a copy, away from your live agent.
- Tested against example cases and live probes.
- Reviewed by a separate reviewer step, which can stop it.
- Released with a backup and a rollback ready.
- Recorded so you can see what changed and when.
What we have, and what we do not have yet.
- Hosted in the US, in Phoenix, Arizona.
- Backups tested: 60 of 60 restores verified.
- A tamper-evident record of actions, archived for six years.
- We do not train models on your data. The list of providers that handle it is on the security page.
- A workspace for each customer. Isolation between customers is still being finished, see the list on the right.
- A SOC 2 report. None exists yet.
- Full-disk encryption on our servers.
- Database isolation between customers on every internal service. It is in place on some and in progress on the rest.
- A signed BAA with us and with the model provider.
- Agents that pass every one of our own attack tests. They do not all pass yet, which is why sending, posting, paying and deleting wait for a person.
Why not just use Claude yourself?
Claude is very good. Using it on its own, or through a desktop tool such as Cowork, is a fine way to get one person more done. Running a business function on it is a different job.
| Where it gets hard | Claude or Cowork on your own | Luminorix |
|---|---|---|
| Across the organisation | A good fit for one person's own work. Sharing it across a team is left to you. | Set up and run for a team, so a new person uses what is already working. |
| Your data | It sits in the account that set it up. Where it goes is yours to manage. | A workspace for each customer, a record of actions, and a plain list of what we do not have yet. |
| What it costs | Model usage adds up with every task, and someone has to watch it. | Model usage is our job to manage. You see the work done, not a usage meter. |
| When nobody is at the desk | Usually runs when a person starts it. | It runs on a schedule and on events, and writes to you in Teams, Telegram or email. |
| Who sets it up and runs it | You do. | We do. A named person on our side answers when something looks wrong. |
Built on Claude, operated by us, embedded in your business. It works when nobody is at the desk, lives in your channels, has a name, a role and a memory, has control built in, and is done for you.
Ask us the hard questions.
Bring your security questionnaire. A founder answers on the call, and what we cannot answer yet we say so.
- A short call with a founder.
- We ask what takes your week before we show anything.
- A person on your team approves or sends. Every action is on the record.