เลิกเดาว่า 'เธอเข้าใจไหม' — ตารางงานบ้านพร้อมรูปภาพ ภาษาแม่ของพี่เลี้ยง ครอบครัวเช็คได้ว่าเสร็จจริง

Source signals

• SIG-20260426-sea-mainland-family-logistics-helper-task-coordination: "Most mismatches begin with vague scope ... Thai caregivers may avoid direct disagreement to preserve harmony" — RECURRING
• SIG-20260426-sea-mainland-family-logistics-helper-rotation-leave: "Need someone for Saturday evening / Agencies rarely handle last-minute" — caregiver-backup-as-a-service overlap
• SIG-20260426-sea-mainland-family-logistics-meal-rotation-multigen: "ขอคำแนะนำ / กินอะไรดี" — multi-gen meal coordination conflicts (helper executes)

What we ship

**Core features (MVP)**

• "Family Playbook" library: family routines (morning, school, dinner, weekend) with photo + pictogram cards; parents pick templates or edit their own
• Each task is shown to the helper in her native language (Burmese / Tagalog / Vietnamese / TH) — auto-translation via Claude with culturally aware prompts (Buddhist context for Thai, Catholic for Filipino helpers)
• Photo-confirm: helper taps "✅ done" and snaps a photo (lunch on table / uniform folded / child picked up); photo lands in family feed
• Kreng-jai detector: AI reads helper confirmation language ("ค่ะ ลองดูแล้ว", "อาจจะ") — if uncertainty pattern, flags "helper might be too shy to say she can't" + suggested private phrasing in her language
• Weekly retro: Sunday evening auto-report "done / skipped / overdue" + suggestions; helper and family each see their version in native languages
• Helper LINE bot: helper doesn't need an app — LINE bot pushes morning tasks, she taps done + photo via bot

**Primary user flow**

1. TH/EN landing — family signs up, picks "I have a helper" (shows interface language for her)

2. Family imports Family Playbook (ready-made templates: "Bangkok Anuban schoolchild", "Bangkok tutor week", "mixed expat household")

3. Helper invited via QR — her phone opens LINE bot in her native language

4. Week 1: family reviews daily summary — where things broke, where kreng-jai signals fire

5. Pricing trigger after 2 weeks: 399 THB/mo (1 helper); 599 THB/mo (multi-helper / nanny+driver+cook)

**What it looks like**

• **Landing**: hero "หยุดเดาว่าพี่เลี้ยงเข้าใจไหม" (Stop guessing if helper understood), split screenshot — confused mom in LINE on left, clean Burmese checklist with ticks on right; CTA "เริ่มฟรี 14 วัน"
• **Main screen (family)**: day timeline; each task with status + photo proof + helper interface language indicator; separate "Disagreement signals" tab with kreng-jai flags
• **Main screen (helper, in LINE)**: "🌅 ยินดีต้อนรับเช้าวันจันทร์ — งานวันนี้ 7 อย่าง" (or Burmese "မင်္ဂလာ မနက်..."), tap for details
• **Output**: evening weekly retro in LINE to parents: "This week: ✅ 24/28 tasks • ⚠️ 3 missed reading-writing — possibly worth a private chat • 💚 helper polite but looked tired Wed"

**MVP build plan (14 days)**

• Day 1-3: backend User+Family+Helper+Routine+Task+PhotoProof; multilingual i18n core (TH/EN/MM/TL/VI/Khmer)
• Day 4-6: Family Playbook templates + builder UI; LINE bot for helper (multilingual deeplink)
• Day 7-9: Claude translation pipeline + kreng-jai detection prompts; photo upload to R2
• Day 10-11: weekly retro generator, push notifications, family dashboard
• Day 12-13: PromptPay via Omise, TH+EN landing
• Day 14: deploy, Pantip family-helper forum + BKK Expat Mums Facebook launch

Reasoning

The pain is documented in three SIGs (helper-task-coordination — top-tier RECURRING, helper-rotation-leave, meal-rotation-multigen) and is actively addressed by commercial nanny agencies (FamBear, BKK Nanny, Kiidu) — but they sell people, not a coordination layer. The vacuum: helper is already hired; language barrier and kreng-jai persist. None of the existing family-app products (Cozi, Famly, OurHome) work with Southeast-Asian languages or understand the kreng-jai pattern.

Stack-fit is ideal: PWA for the family + LINE bot for the helper (Burmese / Filipino domestic workers in Thailand use LINE en masse). Claude TH↔Burmese/Tagalog translation is a standard API call. Photo-proof upload is R2 + simple URL. 14 days — longer because of multilingual i18n + LINE bot but buildable.

Recurring revenue is high — helpers stay in families 1-3 years; the product is glued to the family's operational rhythm.

Quick competitive read

• **Cozi / OurHome / Famly** — generic family planners, English-first, no Burmese/Tagalog, no helper-specific workflow.
• **BKK Nanny / Kiidu / FamBear** — agencies; sell people, no SaaS tool post-hire.
• **Helper Choice (HK)** — HK-only helper marketplace (placement); no daily task ops.
• **Maxhomie (PH)** — Filipino domestic-worker training, content not tool.
• **WhatsApp manual translation** — family workaround; unstructured, no kreng-jai detection.
• **White space**: bilingual helper-OS with photo-proof + kreng-jai detection. Competitive pressure 8/10.

Initial pricing hypothesis

399 THB/mo Solo-helper (1 helper, 1-2 kids) — sits between solo personal-money apps (99-149) and full Pro (599+); 599 THB/mo Multi-helper (nanny + driver + cook). 14-day free trial no card. Anchor: family pays helper 22-35K/mo; 399 THB is ~1.5% of that for clarity and fewer screw-ups.

Distribution hypothesis

1. Facebook BKK Expat Mums (35K), Phuket Family Network, Chiang Mai Mums — soft launch with product-trial gift

2. Pantip Family / Domestic-worker thread — "build-in-public" focusing on kreng-jai pattern (intriguing to Thai readers)

3. TH Twitter — short before/after helper reels

4. Partnership with FamBear / BKK Nanny: agency offers as add-on at placement ("free 1 month with FamBear placement")

5. Burmese / Filipino worker community LINE OpenChat — soft pitch (helpers recommend to families)