TLDR
Wealth Inc is a fictional, self-initiated concept for advised clients. I designed and built a working prototype around occasional attention and accessible financial detail.
01
Asking for attention
The app asks for attention, then gets out of the way.
A compact banner opens a notification sheet. Recommendations expand into allocation comparisons, risk and the adviser’s reasoning; meeting confirmations need less detail. Counts stay consistent across the inbox, banner and sheet. Nothing that needs a decision can be swiped away: Not now keeps it, and anything that closes it sends the adviser an answer. Resolve everything and the banner disappears. I built the behaviour in a working prototype because a still image cannot show whether answering, deferring and returning hold together.
02
Keeping the detail available
Financial detail stays one tap away, with the adviser close by.
The home screen offers a snapshot; performance, assets, debts and protection live underneath it. Messages share one filtered inbox, with account-linked threads for replies. I left charts, market news and engagement mechanics off the home screen because a delegating client has not asked for a trading terminal. That leaves room for their adviser, one tap from a message or call.
03
Giving colour a job
Colour separates money from actions in both themes.
Red is kept for a real loss: a return or a price that fell. Debts, costs and planned reductions stay neutral, unlike my FNZ lo-fi, because a red mortgage on every screen teaches people to ignore red. Green marks wealth and gains, amber a gap that needs attention, blue controls and active states. Percentages and bar lengths also show allocation changes, so direction never rests on colour alone. Light mode needed new values, checked against contrast targets, with the same meanings. Tabular numerals align figures; smaller currency symbols and pence let the amount read first.
04
Testing the central trade-off
The first test is whether calm hides something urgent.
Summarising notifications could hide something time-sensitive, so that is what I would test first: time-to-action against a conventional stacked pattern, before trusting the calm. Lovable let me render and iterate quickly; the attention model, colour rules and quality calls were mine. Motion comes later. The question that matters is whether the attention model works for real clients.