Platoplan

The Problem

Meal planning apps assume American or European kitchens. For a Filipino household, this means constant friction — suggesting ingredients you can't find at the palengke, macros calculated from USDA data that ignores sampalukan or ampalaya, and meal templates that don't match how Filipinos actually eat (rice-first, protein as accompaniment, one ulam plus sabaw).

PlatoPlan starts from the Filipino kitchen. Weekly plans built around what's available at the local wet market. Macros grounded in real Filipino portion sizes. Substitutions that assume you have a rice cooker and a kawali, not a sous vide.

Stack Decision

React Native + Expo for cross-platform reach without native tooling overhead. My work PC doesn't run Xcode, and Filipino target users are ~70% Android — Expo means I can dev on Windows, test on an actual Android phone, and ship to Google Play without platform lock-in.

State management via Zustand for client state, with TanStack Query for async. SQLite via Expo's local module for offline meal history. Gemini API for recipe suggestions because its Tagalog handling is noticeably better than GPT-4's when I tested with sinigang and kare-kare queries.

Data Model Challenges

Filipino food data is sparse online. I combined USDA FoodData Central for Western ingredients with a hand-curated JSON of ~200 common Filipino ingredients — rice types, fish, local vegetables, regional proteins. Accurate macros for kangkong or galunggong required reading Philippine agricultural research papers and normalizing into the USDA format.

AdMob Integration

Monetization is AdMob banner + interstitial. Chose banner for passive revenue, interstitial between weekly plan generation (the only high-engagement moment that feels appropriate for an ad break). Negotiating between user experience and ad density is an ongoing calibration — I've iterated placement three times based on retention data.

What I Learned

Shipping a mobile app on your own forces every cross-cutting decision at once. What I defer in a React web project — push notifications, offline sync, crash reporting, build signing — are Day One concerns in React Native.

EAS Build was the biggest surprise. Coming from web, I expected npm run build to just work; Native builds require a cloud build service, credentials management, and a 15–30 minute feedback loop. EAS handles it well, but the mental shift from "instant rebuild on file save" to "commit, wait, test on device" took a week to internalize.

What's Next

Closed beta through AdMob test users. Filipino-first copy audit (currently English-dominant with scattered Tagalog). Adding a family-sized meal plan option — the current model optimizes for 1–2 people, but most of my target users cook for 4–6.

PlatoPlan is open to feedback from Filipino devs — DM me on GitHub if you want test builds.