All selected work

Mobile / Local-first software / Project notes

FitFlow

A wardrobe that knows what is ready. A local-first wardrobe app that considers laundry, weather, context, and recent wear when suggesting outfits.

FlutterDartSQLiteOpen-Meteo
FitFlowSystem overview
01Local wardrobe
02Availability filter
03Weather + context
04Outfit suggestion
Availability before scoringArchitecture diagram

The problem

An outfit recommendation is only useful if its clothes are available. FitFlow connects recommendations to the actual state of a wardrobe, including items in the wash or still drying.

How the system works

A local SQLite wardrobe tracks clothing and outfit logs. A deterministic availability filter runs before a weighted recommendation score considers weather compatibility, context, and recent wear.

A key engineering decision

The core recommendation flow uses deterministic rules. Optional Gemini support parses style notes, while the local wardrobe and laundry state remain the foundation of the experience.

Implemented scope

Laundry-state tracking, weather-aware outfit recommendations, optional Gemini style-note parsing, and JSON import/export bring wardrobe availability and daily context into a Flutter app.

Scope & limitations

Core recommendations work from local wardrobe state and deterministic scoring. Gemini style-note parsing is optional; RevenueCat subscription integration is a work in progress.

Next project

Salaahkar

Explore case study