Plutus

Two everyday money problems. People know roughly what they should spend, but lose track of small daily purchases: halfway through a ₹30,000 month they're at ₹22,000 and don't know it. And money between friends, like loans, splits and partial repayments, lives in WhatsApp threads and memory.

The obvious answer is a full finance app: connect your bank, add an AI advisor, forecast the month. But neither problem is about missing data. Both are about not looking.

So the product had to be fast enough to use every day and simple enough to trust at a glance. That ruled out most of what a finance app usually adds.

How strong is this? It came from our own habits and the people around us, not a study. The first-week logging metric below is the test.

An offline app you can install on your phone. Adding an expense takes seconds: the amount field is ready when the sheet opens, and the last category is preselected. Home shows this month's spending against your budget and changes colour as you near the limit. A People tab keeps money owed to you and money you owe clearly apart, with full repayment history. Budgets, goals and recurring expenses round it out.

Kaustubh designed the app and ran the build. I owned the product: the problem, the scope, what we cut, and how we'd measure it.

• Bank syncIt would make the numbers effortless, and it would also mean nobody looks at them. Typing an expense is the moment you notice it.
• AI advisor and forecastingEvery number in Plutus is plain arithmetic you can check yourself. A forecast you can't verify is a number people learn to ignore.
• Accounts and a serverNo sign-up. Everything stays on the phone, so the first expense can be logged in the first minute.

• Home leads with what you've spent, not what's left, because spent is the number that moves when you act.
• The safe daily allowance is rounded down, never up. Rounding up would put someone over budget if they followed it exactly.
• With no server there's no sync between devices. Export and import cover backups for now.

Package it for the Play Store (it was built to wrap without a rewrite) and measure the first-week logging rate with real users.