"AI-powered" can mean almost anything, including things you would never want near your account. So here is exactly where the line is at inde ONE: the agent proposes, you approve, and nothing moves until you say so.
What the agent does
It does the parts that are slow and easy to get wrong. It reads across your holdings and the market, pulls fundamentals, technicals, news and the calendar, and turns "is this overlapping what I already own?" into a sourced, plain-English answer. It can draft a basket from your thesis, flag concentration and FX, and lay out a defined-risk structure.
What you do
You decide. Every recommendation comes with the reasoning and the risk attached, and every action waits for one explicit, scoped confirmation. The agent does not place orders on its own, does not rebalance on its own, and does not move money on its own.
Why it is built this way
Because the goal is to make you a better investor, not to replace your judgement with a black box. An agent that explains its homework and then hands you the decision is more useful, and more trustworthy, than one that quietly acts.
The agent does the heavy lifting and the homework. You stay in control and always approve.
When you do approve, execution runs through your own connected broker, fully audited. See the app.