Request a demo
Platform Overview Finance Sales and CRM Procurement and Inventory Projects and Service People and Payroll Ask Boongon Boongon BusinessBoongon Enterprise
Solutions Construction Facilities Management Service and Repair Hospitality Distribution
Why Boongon Why Boongon Security Integrations Move to Boongon Implementation Compliance
Resources Blog Help Roadmap Changelog Status
Company About Contact
Pricing Sign in
Responsible AI

An assistant that inherits your controls rather than sitting beside them.

Most AI features are bolted onto a product and told to behave. This one runs inside it, under your session, against the same rules as a person. Here is exactly where the line is, including the place it is thinner than you might assume.

The boundary

Permissions are not something the assistant is asked to respect.

An instruction in a prompt is a request. The boundary here is the database: queries run under the caller's own session, so row-level security applies to the assistant exactly as it applies to the person using it.

Asked

“Who owes me money right now?”

Overdue across 3 invoices
₱207,065.60
Oldest
Slate Vista Builders

Worked out in ordinary code from your own records, then explained

Then: “chase the oldest one”

  1. Reminder drafted in full Not sent
  2. Your approval rule applies Unchanged
  3. Who asked, and who agreed Recorded

It cannot post to your books on its own, or reach a record your permissions exclude

Process illustration · demonstration data A change drafted but not sent, and the approval rule it still has to meet.

Two layers, and only one of them always holds

Row-level security is the boundary. Above it sits a topic policy that can narrow things further — payroll and banking restricted to an owner, for instance — applied on the server, below the model layer, so a request crafted by hand with a valid token meets the same rules as one typed into the chat.

  • Assistance inside your permissions The assistant reads through your own access, so it cannot surface a record you are not allowed to see.
  • Ask Boongon Ask about your records in plain language, and have work drafted for you. It proposes; you review and approve.
  • Tamper-evident audit trail Every recorded action is hash-chained to the one before it, and a check inside the product reports whether the chain is intact.
Where that second layer is thinner than it looks

The topic policy is looked up per request. If that lookup fails, the request proceeds without topic restrictions rather than being refused — it falls back to row-level security, which still holds. So the honest description is: permissions are enforced in the database and always apply; topic restrictions are a second layer that narrows further, and it is not certain to be applied on every request. We would rather write that here than have your security reviewer find it.

Where the numbers come from

The model does not do the arithmetic.

When you ask what you invoiced last month, the figure is computed in ordinary application code from posted records — the same code behind the report pages — and the answer quotes it. A language model is good at understanding the question and explaining the result. It is the wrong tool for adding up a ledger, and it is not asked to.

This is why a figure the assistant quotes matches the report you would have run yourself. They come from the same place.

What the model does

Interprets what you asked for, decides which records to fetch, and puts the answer into words.

What ordinary code does

Fetches, filters, calculates, validates, applies approval rules and posts. Everything with a number in it.

No accuracy figure

We publish no accuracy statistic for the assistant, because we have not measured one. A number we could not stand behind would be worse than none, and the useful question is not how often it is right but what it is able to do when it is wrong. The answer to that is on this page.

Refusals

Five things it cannot do, and why.

These are not settings you have to find and switch on. They are properties of where the assistant runs.

Your own AI tools

If you would rather use your own assistant, connect it read-only.

There is a scoped connection that lets an AI tool you already use query one company's records. It is read-only by construction, not by configuration: there is no write path behind it to switch on.

Keys are per company, expiring and revocable, every call is logged, and the key stops working when the person it belongs to loses access.

See what connects, and its real status
  • Read-only connector for your own AI tools A scoped, read-only connection that lets an approved AI tool query one company’s records. It has no write path.
  • Access enforced in the database Permission rules live with the data, so they hold whether a request comes from the screen or anywhere else.
  • Multi-factor authentication Two-step sign-in is required to reach a workspace, and the requirement is applied by the database rather than only by the screen.

Bring your AI questionnaire.

If your security review has a section on AI, send it before the call rather than after. The answers that are 'no' are quicker to give than to discover in week three.