Framing the problem
In the first weeks with a newborn, the questions are simple and relentless: when did she last feed, which side, how long did she sleep, is this vaccine due, is this weight okay, what did I want to ask the doctor? The apps I tried in those weeks all seemed designed for a different parent — one with two free hands, good light, and the patience to fill in forms. They demanded data and returned dashboards.
What a sleep-deprived parent actually needs is narrower and kinder: log this instant with one thumb, tell me what's coming next, and tell me my baby is okay — in words, not charts. Dewdrop is my attempt at that app.
Design question: How might a baby-health app serve a parent at 3 a.m. — one-handed, half-asleep, and quietly anxious — without adding to the anxiety?
The user was me
Me, at week three
Every design decision in Dewdrop traces back to a specific memory from my own first months as a mother: forgetting which side the last feed ended on; the immunisation schedule living as a photocopied table folded into a diary; arriving at the pediatrician's clinic and discovering that every question I'd had at 3 a.m. had evaporated by 10 a.m.; and midnight searches about growth percentiles that made a healthy baby feel like a problem to solve.
A sample of one isn't research — but it's the most honest starting point I've ever designed from. Every screen had to survive a test against a real memory. In a real engagement, I'd validate all of it with other parents (see "What I'd do next").
Process & key design areas
Logging: one thumb, two taps, in the dark
Feed, sleep, and diaper events are logged in at most two taps from a bottom-anchored home screen — large targets in the thumb's natural arc, because the other arm is holding a baby. The app remembers context: it pre-selects the likely side for the next feed and defaults timestamps to "just now," with one-tap chips to backfill ("started ~20 min ago") when you inevitably log late. The interface is dim-light-first: a near-black theme at night so the phone doesn't wake the person you just spent forty minutes putting down.
Today: a rhythm strip, not a dashboard
The day is drawn as a single gentle horizontal strip — feeds, sleeps, and wake windows in sequence — because what a parent reads from it is rhythm, not totals. Above it, one line in plain words: "She's usually ready to sleep around now." Counts and durations exist, but a level down; nothing on the default screen keeps score.
Vaccination & checkups: the calendar carries the load
The standard immunisation schedule (in India, the IAP schedule the pediatrician follows) is built in from the baby's date of birth — no manual setup. Each vaccine gets a plain-language card: what it protects against, what to expect afterwards, and a short prep checklist for the visit. Reminders arrive ahead of the due window, phrased as care rather than compliance — "Her 10-week vaccines are coming up" — and a photo of the clinic's paper record card can be attached to each entry, because the paper card is still the source of truth at most Indian clinics.
Growth without anxiety: bands, not verdicts
Growth is where well-meaning apps quietly frighten parents. Dewdrop plots weight, length, and head circumference against the WHO growth-standard curves — but as a soft band, not a single line to be above or below. The default reading is a sentence, not a number: "Following her curve steadily." The app never issues verdicts; anything unusual is framed as "worth mentioning at the next visit," because interpreting growth is the pediatrician's job, not the app's.
Visit notes: the questions you had at 3 a.m., ready at 10 a.m.
Any moment of worry can be captured as a one-line question the instant it occurs — typed half-asleep or dictated — and it files itself under the next appointment. Before a checkup, Visit Mode compiles a single card: what changed since the last visit (feeds, sleep, growth, anything flagged), the open questions, and the vaccines due — designed to be read aloud or simply handed over in the consultation's first minute.
Key decisions, summarized
- Two taps over rich data entry: a log that's 90% complete and actually kept beats a perfect one that's abandoned by week two.
- Words before charts: every screen leads with a plain-language sentence; numbers and curves are one level down.
- No streaks, no scores, no comparisons to other babies — a newborn's first year is not a performance.
- Reminders phrased as care, not compliance — the app is on the parent's side, never auditing them.
- Privacy as a feature: an infant's health data is among the most sensitive a family has. The concept is local-first, with explicit export to share with a doctor — and nothing social.
What I'd do next
The honest next step is to grow beyond a sample of one: interviews and diary studies with other new parents — fathers and grandparents as caregivers, parents of twins, NICU parents whose tracking needs are clinical, not gentle. Then usability tests of the logging flow under realistic conditions (one hand, dim light, interrupted mid-task), a content-accuracy review of the vaccination material with a pediatrician, and comprehension testing of the growth band — the screen where my "sentence first, chart second" hypothesis most needs evidence.
A personal footnote: Dewdrop is the one project on this site I designed for myself. Our baby arrived in August 2025, and in the months that followed I kept sketching the app I wished was already in my hand — designing the experience I wished I'd had. The plan was to build it the way we built Biller — my design, my husband's code — but he was heads-down in his own work as a software engineer at Google, so we set the build aside. Dewdrop stayed what you see here: a design concept. The newborn nights it was designed for have passed; the care it was designed with hasn't.