Properties Inventory Setup
Configure the searchable property catalog. Populate specifications, rates, mapping information, and visual media to feed your WhatsApp conversational agent.
Unified Property Repository
The Properties Inventory dashboard operates as the single source of truth for all listings searchable by the WhatsApp AI. It stores geographical coordinates, building categories, layout specs, pricing metrics, and registration credentials.
When database rows are added or modified, they sync in real time with the chatbot's tool calling framework, allowing immediate, context-aware properties query handling without needing model retraining or configuration adjustments.
1. Inventory Metrics
The top-level properties interface compiles inventory statistics to monitor distribution:
Total Listings
Sum total of all active properties in database.
Active Availability
Listings whose state is explicitly set to available.
Outright Purchase
Listings flagged under listing type sale.
Leasing & Renting
Listings matching rent or lease listing types.
2. Property Schema Specifications
Each listing is stored using precise parameters to enable the AI agent to filter matches accurately. The core database fields include:
| Property Field | Database Column | Type | Technical Details |
|---|---|---|---|
| Title | title | Text (Required) | Public title of the listing (e.g. 3 BHK Luxury Seafront Villa). |
| Property Type | property_type | Text | Classification: apartment, villa, penthouse, plot, commercial. |
| Listing Type | listing_type | Enum | Allowed tags: sale, rent, lease. |
| Bedrooms / Baths | bhk / bathrooms | Integer | Total counts of bedrooms (BHK layout marker) and bathroom units. |
| Area size | area_sqft | Numeric | Total built-up or carpet area size in square feet. |
| Pricing Metrics | pricerentdepositmaintenance | Numeric | Financial values mapping to outright purchase price, monthly lease/rent, initial deposits, and recurring building maintenance fees. |
| Geographics | citylocalityaddress | Text | City name, neighborhood/sector locality, and literal street address. Also captures Google Maps coordinates. |
| Legal & Status | rera_idpossession_statusavailability_status | Text / Enum | Official RERA registration numbers, possession timelines (ready to move/under construction), and transaction flags (available, sold, rented, or inactive). |
| Media Assets | image_urlimages | Text / Array | Primary property thumbnail image path and an array listing additional gallery URLs or walkthrough videos. |
3. Interface Controls & Media Operations
The properties dashboard incorporates layout builders and cloud storage connectivity:
Three-Step Form Dialog
To keep inventory listings pristine, properties are generated through a multi-step modal layout:
- General Details: Main attributes, RERA numbers, BHK layout, description text, and furnishing status.
- Rates & Site: Regional mapping, local landmarks, Google Maps link, monthly maintenance, and pricing.
- Visuals: Media attachments and cloud storage bindings.
Supabase Storage Integration
Primary listing visuals are processed through native drag-and-drop cloud upload controls:
- File streams are sent directly to the Supabase Storage bucket named
property-images. - Upload structures are sandboxed under matching user IDs (
owner_uuid/file_name). - Retrieves a secure public URL automatically, binding it back to the property's
image_urlfield.
Demo Inventory Seeding
To test your WhatsApp properties setup quickly, the dashboard includes a single-click seeding function:
- Clicking Add Demo Properties seeds 3 luxury listings (a Palm Jumeirah Villa in Dubai, a Downtown Manhattan Studio, and a Central London Penthouse).
- Pre-populates sample coordinates, rates, registration tags, and high-quality mock media URLs.
Tool Integration Details
How WhatsApp AI agents parse your property records
When an end-user queries the WhatsApp channel, the LLM utilizes properties database parameters inside search functions. System behaviors are mapped below:
Crucial Information for Prompt Configuration:
- Exact Query Matching: Ensure the listing names and locality columns match local terminology. AI filters search results based on exact substring filters on localities and city names.
- Currency Parsing: The tool supports multiple currencies (
INR,USD,AED,EUR,GBP). The AI agent converts financial numbers and states them clearly to the customer based on the selected setting. - Fallback to On-Request: If listing price fields are left blank, the dashboard outputs On request. The AI agent will detect this state and prompt the client to share details for an agent callback instead of saying the price is zero.