Braza

About

Operations platform for a high-protein meal prep service, turning each client's biometrics into meal plans, pricing and kitchen production.

Role

Design Engineer

Year

2026

Braza customers list with a command bar for creating a customer — Customers: client database with an AI command bar that turns plain-language prompts into customer records and weekly orders.

Overview

Braza was a planned meal prep service for gym-goers and athletes, delivering high-protein meals portioned to each client's body and goals. That level of personalisation breaks down when done manually, with biometric calculations for every client and hours of planning, pricing and kitchen prep every week. The platform runs the whole workflow, combining a macro engine, yield-adjusted production sheets and a bilingual AI command bar for rapid order entry. It was paused before launch given the scale of the physical operations.

Outcomes

Eliminated manual bottlenecks to make personalised meal prep scalable
Automated meal planning and weekly pricing from 8 biometric inputs
Generated yield-adjusted production sheets, cutting hours of manual math per batch
Built an EN/PL interface & AI command bar capable of processing 100+ weekly orders in under an hour

Front-end

Next.js · React · TypeScript · Tailwind · shadcn/ui · next-intl

Back-end

Supabase (Postgres) · Anthropic Claude API · Vercel

Braza add customer form with calculated calories, protein and price — Macro engine: biometrics resolved into calorie targets, protein requirements, portion scaling and pricing.

Braza add order form with meals, delivery and weekly total — New order: weekly meal count priced automatically from the client's biometric requirements.

Braza production sheet for the week, with meals, portions and allergy notes — Production sheet: active orders aggregated into prep lists with portion weights and allergy flags.

Braza meals built from a protein, base and side — Recipes: ingredients composed into protein, base and side components for each meal.

Braza shopping list with raw purchase weights for proteins, bases and sides — Shopping list: raw ingredients needed for the week, calculated from yields and every customer's portions.

Braza Supabase schema for meals, orders, production and the weekly plan — Database schema: Postgres relational model mapping client profiles, biometric state, modular recipes, order lines, and weekly production batches.

Braza customers page in Cursor, with locale routes and Claude Code — Architecture: Next.js App Router with next-intl routing, server-rendered Supabase queries and a single shared library for all macro and pricing calculations.