📊 Project Status Brief
Last updated: 2026-03-30
Pantry
Voice-first household OS for South African families. Updated after every working session.
4
Sprints complete
6
Screens live
76
Backlog stories
3
Active testers
SA
Market focus
🎯 Current focus — Sprint 5 planning
Sprint 4 (budget + household management) is complete. The next build is Sprint 5: Intelligence — recipe engine, smart shopping list, predictive restocking, email order parsing, and alerts. Sprint 5 scope not yet formally kicked off. Two blockers to resolve first: onboarding flow design and monetization model definition.
⚠ Open items requiring attention
Onboarding flow not designed Monetization model undefined Huma tester feedback pending Waitlist not built Pantry Pro: no sprint allocated
What this product is

Pantry is a household operating system for South African families. It starts with grocery and inventory management — tracking what's in the house, what needs to be bought, and what's been spent — and grows into a unified layer for all household coordination: bills, domestic worker tasks, family schedules, and smart home automation.

The core insight: SA households have no single place where food, spend, and domestic coordination live together. They're managed across WhatsApp groups, notes apps, spreadsheets, and memory. Pantry replaces that fragmentation with a passive, voice-first system that works with minimal effort from the household manager.

The product also has a B2B commercial tier: Pantry Pro — the same platform extended for restaurant owners (Sipho Dlamini persona) who need real-time stock visibility, supplier management, food cost tracking, and sensor-driven automation. The infrastructure is shared across both products.

Locked decisions
Strategy
Household OS — not grocery-only
Scope confirmed. Bills, smart home, family coordination, home maintenance all in scope. Grocery is the entry point, not the ceiling.
Strategy
SA-first market
R420B grocery market, 6.8M target households. Load-shedding, Woolies/Checkers/Makro ecosystem, WhatsApp-first behaviour all built into the design.
Strategy
Platform: HOME + PRO on shared core
Pantry Home (household OS, freemium) and Pantry Pro (restaurant OS, paid) share the same Supabase schema, Edge Functions, and capture infrastructure.
Product
Passive capture first
Receipt scanning (OCR) and barcode scan over manual entry. Automation is the value proposition. Manual entry = churn risk.
Product
Voice-first, multi-surface
Web app now. WhatsApp bot (Phase 2) for Thabo and Grace. Nest Hub voice (Phase 2). Meets users where they already are.
Technical
React + Supabase + Cloudflare Pages
RLS for per-household data isolation (POPIA). Google Cloud Vision for OCR. All live in production. Locked since Sprint 1.
Personas
Lerato
Household manager
Affluent SA family. R25k/month grocery spend. Shops at Woolies, Checkers, Makro. Manages all household members, budget, and domestic staff.
Primary
Thabo
Partner / shopper
Delegated grocery runs. Needs mobile-first shopping list that works in-store on poor signal. Marks items purchased, syncs back.
Active
Grace
Domestic worker
Adds items by voice or WhatsApp. Scans barcodes. Must work without a smartphone app install.
Phase 2
Sipho
Restaurant owner (Pro)
Casual dining, Joburg. Sensors on containers and fridges. Needs live stock view during service. Load-shedding = silent stock loss.
Future ICP
What's built — Sprint 4 complete
Auth + Household setupSign up, sign in, household creation, member invite, roles
Inventory managementCRUD, baseline stock, expiry tracking, search/filter
Receipt scanning (OCR)Google Cloud Vision Edge Function — live in production
Barcode scanningCamera-based item lookup and add
Budget trackingMonthly periods, category and store breakdown
Household managementMember roles, permissions, recurring items
Recipe engineSprint 5 — from current inventory, gap detection
Smart shopping listSprint 5 — auto-generate from recipes + consumption
Low-stock + expiry alertsSprint 5 — notifications when thresholds hit
Email order parsingSprint 5 — Sixty60, Woolies Dash confirmations
Onboarding flowNot designed — critical gap before Sprint 5
WhatsApp bot + IoT sensorsPhase 2 (Grace, Thabo, Melissa's stack)
Live app: pantry-app.tapfumamv.com  ·  Supabase: osmtvniihyxpkadjvyur.supabase.co  ·  8 DB tables · RLS live · 13/13 integration tests passing

Open items — needs a decision or action
🔴
Onboarding flow not designed
No first-use experience. A new Lerato has no guided path from sign-up to first value. Products live or die at this step. Must be scoped before Sprint 5 kicks off.
Critical
🔴
Monetization model undefined
No pricing model, no subscription tiers, no revenue hypothesis. Working hypothesis: household freemium + Pro paid tier. Willingness to pay not yet tested. Define before Sprint 6.
Critical
🟠
Huma tester feedback pending
4th tester (single woman profile) not yet onboarded. Her feedback validates Phase 2 scope and may surface requirements not in the backlog.
Important
🟠
No user acquisition plan
Waitlist not built. Beta programme not defined. Currently 3 testers. The next milestone is 50 real households.
Important
🟠
Pantry Pro — no sprint allocated
Restaurant OS strategy defined, stories logged (PRO-001–012), persona documented (Sipho). No build sprint assigned. Prioritization decision needed.
Important
Reading guide — where to go from here