Stay ahead of repairs, rent issues, and resident escalations.
Lette AI gives property managers one place to review resident, landlord, and contractor conversations across email and WhatsApp, with the urgent issues surfaced first and the next step made clear.
| Spot the real emergencies | Unify the channels | Move to action faster |
|---|---|---|
| Detect repeat complaints, legal risk, and operational failures before they disappear in the queue. | Keep email and WhatsApp in one working inbox with shared context. | Reply faster, dispatch sooner, and reduce back-and-forth across your team. |
Built for property managers who need to spot what matters now, keep service levels high, and stop important messages from slipping through the cracks.
Property management teams deal with a constant flow of repairs, rent questions, move-in issues, contractor updates, and resident complaints. The real problem is not message volume alone. It is losing context.
A message about a broken fridge can look routine on its own. If it is the third follow-up after a missed repair visit, it is no longer routine. It is an escalation, and by the time someone notices, the resident experience has already worsened.
Current tools fail to connect the dots between:
- Historical Context: Previous failed attempts or "false promises."
- Channel Fragmentation: Vital info split across Email, WhatsApp, and SMS.
- Urgency vs. Noise: Routine inquiries burying legal and safety risks.
Lette AI is a context-aware triage workspace for property managers. It turns raw, messy communication into an organized daily queue so teams can see which conversations need attention first, why they matter, and what to do next.
From one dashboard, teams can review ongoing enquiries, scan messages by channel, and open full thread history without bouncing between inboxes.
- Prioritize faster: See repeat complaints, sensitive issues, and emerging risks before they become service failures.
- Work from one view: Keep resident, landlord, and contractor communication visible in the same workflow.
- Respond with confidence: Use AI-assisted drafts and routing suggestions to act quickly without losing context.
- Intelligent Classification: Automatically sorts conversations into Maintenance, Rent, Leasing, Legal, and other operational categories.
- Repeat Complaint Detection: Spots when a resident is chasing the same issue again and escalates the thread accordingly.
- Action Suggestions: Generate a reply draft or identify the next operational handoff, such as a contractor dispatch.
- Sentiment Heatmapping: Highlight tense or deteriorating conversations before they become formal complaints.
| Area | What it helps with |
|---|---|
| Ongoing enquiries | See active issues by type and understand where pressure is rising |
| Thread list | Review incoming messages by channel, sender, preview, and timing |
| Thread detail | Open the full conversation and act with the right context |
Data is loaded from the Messages API (GET /api/messages), which reads new_data.json (emails) and merges with incoming WhatsApp messages from the Twilio webhook. The dashboard polls every 8 seconds so new WhatsApp messages appear automatically.
Play the emergency scroller demo (MP4)
| Inbox | Message detail |
|---|---|
 |
 |
| Categories | Sorting modal |
|---|---|
 |
 |
When someone sends a message to your Twilio WhatsApp number, it is stored and shown in the triage inbox.
-
Twilio setup
- Sign up at Twilio and activate the Twilio Sandbox for WhatsApp (or use your own WhatsApp Business sender).
- Join the Sandbox by sending the required message to the Sandbox number from your phone.
-
Webhook URL
- Your app must be reachable from the internet so Twilio can POST to it.
- Local dev: use ngrok, e.g.
ngrok http 3000, then use the HTTPS URL as below. - In Twilio Console → Try WhatsApp → Sandbox settings, set When a message comes in to:
https://YOUR_HOST/api/webhooks/whatsapp
(e.g.https://abc123.ngrok.io/api/webhooks/whatsappfor local dev).
-
Flow
- User sends a WhatsApp message to your Twilio number → Twilio POSTs to
POST /api/webhooks/whatsapp→ the message is appended todata/incoming-messages.json→ the dashboard fetchesGET /api/messages(and polls every 8s), so the new message appears in the inbox.
- User sends a WhatsApp message to your Twilio number → Twilio POSTs to
-
Production
- File-based storage (
data/incoming-messages.json) is used for simplicity. For production, replace this with a database and update the webhook andGET /api/messagesto read/write from it.
- File-based storage (
Docs: Twilio Docs, WhatsApp quickstart.
# install
npm install
# run dev server
npm run devOpen http://localhost:3000.
# build for production
npm run build
npm start- Next.js 14 (App Router)
- React 18 + TypeScript
- CSS (no UI framework) — layout and bento grid
├── app/
│ ├── api/
│ │ ├── messages/route.ts # GET: merged messages (static + WhatsApp)
│ │ └── webhooks/whatsapp/route.ts # POST: Twilio WhatsApp webhook
│ ├── globals.css # global styles
│ ├── layout.tsx # root layout
│ └── page.tsx # main triage inbox page
├── data/
│ └── incoming-messages.json # WhatsApp messages from webhook
├── new_data.json # email dataset (metadata + emails[])
├── components/
│ ├── ChannelBlock.tsx
│ ├── Header.tsx
│ ├── OngoingEnquiriesHierarchy.tsx
│ ├── Sidebar.tsx
│ ├── ThreadCard.tsx
│ ├── ThreadDetail.tsx
│ ├── ThreadList.tsx
│ ├── icons/ # EnvelopeIcon, WhatsAppIcon
│ └── ui/ # BentoGrid, BentoCard
├── lib/
│ └── utils.ts
├── types/
│ └── message.ts # Message, Channel, UrgencyLevel
└── Lette_AI_Triage_Inbox_PRD.docx
Email data lives in new_data.json at the project root:
{
"metadata": { "dataset_version": "1.1", "total_entries": 100, "properties": [...], "sender_types": [...], "urgency_levels": [...] },
"emails": [
{
"id": "email_001",
"thread_id": "thread_001",
"timestamp": "2026-03-06T08:12:00Z",
"from": { "name": "...", "email": "...", "type": "tenant", "unit": "Apt 14B", "property_id": "prop_001" },
"to": "citynorth@manageco.ie",
"subject": "URGENT - Water leaking...",
"body": "Hi,\n\n...",
"attachments": ["file.jpg"]
}
]
}The API maps each emails[] entry into the internal message shape and merges with WhatsApp messages from data/incoming-messages.json.
| Command | Description |
|---|---|
npm run dev |
Start Next.js dev server |
npm run build |
Production build |
npm run start |
Run production server |
npm run lint |
Run Next.js lint |
*Built for the Give(a)go Hackathon - Dublin 2026*