How it handles company data.
Six guarantees, in plain words. Every one is visible in the run log.
Every tool is a switch you hold
A connected app lists every tool it offers, each with a switch, in your own account. Nothing a run does can turn one back on. If there is a tool you never want used, such as a send, a delete or a refund, switch it off. It stays off whatever anyone later configures.
Review any app, when you want it
The control is the switch: each person connects their own account and decides, tool by tool, what Ara may do with it. Nothing outside that ever runs. On top of that, a workspace admin can put any app under review, so every change in it is previewed and waits for a member's approval, or block an app outright. Review is opt-in, so approval queues stay short enough to be read. Every action is written to the run record either way.
Works in its own sandbox
One isolated machine per workspace, no shared filesystem, outbound hosts pinned to an allowlist. Nothing the model asks for runs on our servers.
Never sees your keys
Tokens are encrypted at rest and stay on the API host. The sandbox holds a short-lived, run-scoped credential and nothing else.
Treats the web as data, not orders
Pages, files, tickets and emails cannot give it instructions: their content is data to report, never a request to act on. A run started by an inbound event is told so explicitly, and told to draft rather than act when the payload seems to ask for something outward-facing.
Your data is never training data
Customer data is used only to fulfil that request. It is not used to train AI models, ours or anyone's.
Where it can be wrong
Like any model, Ara can be wrong. That is why work comes back with the steps it took and a file you can check, why every run is on record, and why you decide tool by tool what it can reach at all.
Your team's first AI employee. Starting today.
Free to start, no card. Connect the messenger your team already uses, hand over one job, and the finished work comes back the same day.