Bunny is a client manager for one independent financial advisor. It holds the client book, and it remembers every call, meeting and deadline so the advisor doesn't have to.
Started from a real annoyance, not a brief. Designed, built and shipped end to end with a co-founder.
39 screens in Kotlin and Jetpack Compose, built around one job: never let the advisor forget a client.
Quarterly portfolio reports and half-yearly reviews, scheduled for the whole financial year automatically.
Testing found advisors who also do tax. Instead of one bloated app, the app reshapes itself by profile.
My own mutual-fund and tax advisor missed a deadline, and never told me it was coming.
It kept happening. I was the one messaging him about my portfolio review. A client chasing his advisor for the service he was already paying for.
He is not careless. He has a few hundred clients and no system — nothing to hold a client's details, nothing to tell him who to call this week, nothing to schedule the reviews he had promised.
The work was not the problem. Remembering the work was the problem.
Deciding who this was not for shaped everything after it.
Teams, roles, permissions, dashboards for a manager. That is a different product with a different sales cycle, and it is not the person who was failing me.
Enough clients that memory fails. Few enough that they know every name. No back office, no assistant, no software budget — just a phone.
Because the user is one person on a phone, the app never asks anyone to log into a desktop, and every important action finishes in a couple of taps.
No tax. No pipeline. No reports. Reminders for meetings, and wishes on the days that matter.
Splash · client manager, financial companion
Onboarding · the promise, in four lines
Onboarding · what it does for you
Start from the phone book the advisor already has, not an empty table.
Name, mobile, email, address, birthday, notes, photo, tag.
The reason the app exists — nothing gets promised and then forgotten.
The advisor's relationship work, made one tap instead of a diary entry.
Asking a forgetful person to remember to set reminders solves nothing. So the app sets them.
Auto-generate · one tap, one year
Add a client, press one button, and the year is scheduled:
Portfolio report on the 4th day after each quarter end — 4 July, 4 October, 4 January, 4 April.
Review meeting twice a year — 15 June, 15 December.
Existing reminders are skipped, so pressing it again never creates duplicates.
This is the whole product in one dialog. The advisor's promises become dated entries the moment the client exists.
Add client · KYC alert, 60 days before a 3-year expiry
Client · six reminders, already dated
Manual reminder · for anything the rules miss
Meeting · schedule, then send the invite
Meetings carry mode, link, minutes and a follow-up status — so the next reminder can be created out of what actually happened in the last meeting.
Many of the advisors testing the app were also handling their clients' tax — GST and income tax — with no tool for it.
The same person is a fund advisor to a client on Monday and their tax consultant in July. Two jobs, one contact list, two sets of deadlines.
Giving everyone everything would have buried a pure fund advisor under GSTR-9C, and a pure tax consultant under SIP anniversaries.
“What best describes your practice?” Financial advisor, tax consultant, or both.
Two tabs stay fixed — Home and Clients. Two change. The fifth is an overflow hub, so nothing is lost, it is just not in the way.
Changeable later in Settings, because people's practices change.
First launch · the question that reshapes the app
Home · GSTR-1 in 13 days, advance tax in 17
More · grouped by job, not by menu order
Activity · meetings and reminders together
Deadlines are counted, not just listed — “13 days”, “17 days”, “22 days”. A date alone is something to interpret; a countdown is something to act on.
This app opens in front of a client. It should look like a professional's tool.
A single emerald marks the primary action on any screen. Status uses amber and red only on the thing that is late.
Jetpack Compose with Material 3, so the app inherits Android behaviour — dynamic type, dark theme, back gestures, accessibility — instead of reimplementing it badly.
Rather than argue about taste, I built the same home screen four ways.
Being the founder means nobody else will tell you the pricing is wrong.
But birthday alerts, SIP anniversaries and KYC warnings only fire with volume. A free user would never see the product work.
This app is opened in front of clients. An interstitial mid-KYC damages the advisor's credibility, and mine.
To an advisor earning lakhs in trail commission, ₹99 is not cheap — it is invisible, and it signals a toy.
Written against my own decisions, from reading the shipped code rather than the plan. Owning the product includes being willing to be wrong in public about it.
Nobody asked for this. I was the client being let down, which meant I knew exactly what “working” looked like.
A reminder app that needs you to set reminders has not solved the problem. Generating the year is the feature.
Finding that advisors also do tax added a second half — and the profile system exists only because giving everyone everything would have ruined it.