The "haan le li" problem — designing medicine reminders that actually confirm

Here's how it usually goes. Your parent has three medicines in the morning. You set an alarm on their phone. For the first week, the alarm rings, they take the medicine, they feel good about it. By week three, the alarm is being dismissed half-asleep and the medicine is being taken later, or not at all.

You then switch to calling every morning. This works better, but it costs you the call and it costs your parent a small dignity — being checked up on like a child. After a few weeks, when you ask "did you take your medicine?", the answer is always "haan le li" — yes, I took it. Whether they did or not.

This is the problem we were trying to solve. Not the reminder part — that's easy. The confirmation part.

Why alarms fail

The issue with alarms is that dismissing one becomes automatic over time. The alarm rings, the hand reaches out and dismisses it, the intention to take the medicine is still somewhere in the mind — but the action happens later, or not at all. There's no friction between "dismissed" and "taken." They feel identical in the moment.

Apps with push notifications have the same problem. The notification appears, gets tapped away, and the app expects you to then go into it and tap "taken." That's two steps too many for a habit that's competing with morning chai and the newspaper.

The verbal confirmation insight

What works better is a conversation. Not a notification you dismiss — a message that expects a reply. This replicates the structure of the morning call, but without the call.

FamAI sends a message at the scheduled time: "Subah ki dawai ka time ho gaya — Amlodipine aur Vitamin D. Le li kya?" The parent replies — "le li", "haan", "abhi le raha hoon" — and FamAI logs it. The reply is the confirmation. It takes one tap and two words.

This works because the response creates a small moment of accountability. The parent isn't dismissing a notification. They're answering a question from something that remembers their answer. That's a different psychological trigger.

It doesn't solve the verification problem — we can't know if your parent actually took the medicine or just replied. But it reliably improves engagement compared to passive alarms, which is the right thing to optimize for in a population that uses WhatsApp all day.

Why 30 minutes, not 5

When we designed the follow-up, we debated the timing. 5 minutes feels urgent. We chose 30 minutes for a specific reason: mornings are complicated. There's breakfast to make, a grandchild to call, a serial that just started. The parent might have seen the reminder, meant to respond, and genuinely gotten absorbed in something else — not forgotten, just delayed.

A 5-minute follow-up feels like nagging. A 30-minute follow-up feels like a gentle check. The distinction matters. We're building something your parent interacts with willingly every day, not an alarm clock they learn to resent.

We also cap it at one follow-up. If there's still no response, the dose is logged as no-response and we stop. Sending a third or fourth message is not respectful.

The streak design

We added streak tracking — consecutive days where all reminders were responded to — after observing that some parents were motivated by it and frustrated when a streak broke. The weekly summary gives them something to feel good about without turning medicine adherence into a game.

For families who want to track this more carefully, the adherence log is available. But we deliberately chose not to push individual reminder responses to the adult child in real time. That level of surveillance would make this feel like monitoring, not care.

The family loop

This was a deliberate decision: the adult child does not get a notification every time a reminder is sent or confirmed. They see the adherence summary — the week's trend — not a real-time feed. The parent's daily medicine routine is not something that should land in their kid's phone as a ping. That dynamic doesn't feel like care; it feels like supervision.

The right trigger for the kid to get involved is a trend problem, not an individual missed dose. One missed dose happens to everyone. A week of low adherence is the signal that something is wrong.