Amenta is the ARMANET real estate intelligence system — a platform that collects, deduplicates, and analyzes every property listing across every major source in the market, then reads the images using AI vision to extract information that no text-based system can see. The result: a single source of truth for the real estate market, built on data rather than listings.
The Noise Problem
The real estate market generates millions of data points across hundreds of platforms, broker websites, and developer launches. The same apartment appears on five portals with five different prices, described by five different agents, photographed at five different angles. A listing is removed, then reappears as new three weeks later at a lower price. A photo from one property is used to market another. A developer claims a district is booming while the data shows the opposite.
Amenta cuts through all of it. Every listing is collected, every property is identified, every image is analyzed, and every claim is measured against data. What remains is the real market.
Multi-Source Collection
Amenta aggregates listings from every major source: property portals, broker websites, developer launches, classified platforms, and auction announcements. The collection is continuous — not a one-time scrape. New listings are detected within minutes of publication, price changes are captured in real time, and delistings are logged with their timestamps.
Each source is monitored independently. When a listing disappears from one portal but remains on another, Amenta tracks the discrepancy. When a property is listed simultaneously across platforms, Amenta captures every version. Nothing is missed, nothing is assumed.
Property Identity and Deduplication
The core challenge in real estate data is deduplication: the same property listed multiple times across different platforms, described differently, priced differently, and photographed differently. Traditional systems rely on manual matching or basic text comparison — both unreliable.
Amenta uses a multi-signal identity engine that combines: textual analysis (descriptions, addresses, specifications), numerical matching (price, area, rooms, floor), and AI vision (image matching across sources). When five listings across five platforms refer to the same apartment, Amenta creates a single canonical property record with all five versions linked, all price points preserved, and all images associated. The market data becomes clean.
Price Intelligence
For every property, Amenta builds a complete price trajectory: original listing price, every reduction, every increase, every relisting at a different price, and the final sale price where available. The history is visualized as a timeline with each event timestamped and sourced.
Price drop detection goes beyond simple threshold alerts. Amenta distinguishes between: a standard negotiation-driven 2-3% reduction, a motivated-seller 10%+ drop, and a market-driven cluster of reductions across multiple properties in the same district. Each pattern triggers a different signal with a different interpretation.
Area-level price analytics aggregate individual property trajectories into district trends: average price per square meter, median listing duration, price dispersion, and month-over-month and year-over-year changes. These are presented in Arabic with trend lines and district comparisons.
Relisting and Market Behavior
A listing that is removed and re-published as new is a relisting. It is one of the strongest behavioral signals in real estate: it indicates a property that failed to sell, a seller becoming motivated, or a market that is slowing down. Amenta detects relistings automatically by matching property identity across time, even when the description is rewritten and the price is changed.
Days-on-market is calculated from the original first listing, not the relisting date. A property that has been on the market for 180 days — despite appearing as a fresh 3-day listing — is accurately classified as stale inventory, not new supply.
Comparable Property Analysis
For any property in the system, Amenta identifies true comparables: same district, similar size (within configurable tolerance), same property type, similar condition, and similar amenity profile. Comparables are ranked by similarity score and presented alongside the subject property with price, area, price-per-square-meter, and days-on-market.
The comparable analysis enables automated valuation assessment: whether a property is priced above, below, or at market level — with the comparable evidence attached. Banks use this for collateral assessment. Developers use it for launch pricing. Investors use it for opportunity identification.
Area and Market Analytics
District-level intelligence presented in Arabic: inventory levels (active listings, absorbed listings, stale inventory), price trends (average, median, price-per-square-meter), supply dynamics (new listings, new developments, pipeline), and demand indicators (inquiry proxies, listing views where available, days-on-market compression).
Heatmaps visualize price density and activity across districts. Comparative analysis ranks districts by appreciation rate, liquidity, and investment yield. Supply-demand balance indicators flag districts approaching oversupply or shortage.
Risk and Opportunity Signals
Amenta generates actionable signals from the data. Opportunity signals: undervalued properties (priced significantly below comparable median), motivated sellers (relisting with price drops), districts with accelerating appreciation, and pre-launch investment windows. Risk signals: oversupply forming in a district, properties with extended days-on-market, price-decline clusters, and developer inventory concentration.
Each signal includes the underlying data, the comparable evidence, and the confidence level. Your team makes decisions based on data, not opinions.
AI Vision
This is where Amenta separates from every real estate analytics platform in the region. Through the Foblax research partnership, Amenta applies advanced computer vision models to property images — extracting information that text-based systems cannot see.
Image Analysis and Deduplication
Every image in every listing is analyzed. The system detects: the same photo used across different listings (a common practice in misleading marketing), near-duplicate images (same room, different angle), and visually similar properties that may or may not be the same.
When the same photo appears in listings for different properties, Amenta flags the listings as potentially misleading. When near-identical images appear across different platforms, Amenta uses them as additional evidence for property deduplication.
Cross-Source Image Matching
The same apartment photographed by two different agents, published on two different platforms, described with different text. Text matching alone cannot link them. Image matching can. Amenta compares visual features — room layout, window positions, ceiling details, flooring patterns — to confirm that two listings from two sources refer to the same physical property.
This capability is critical for accurate deduplication and for detecting when agents re-list the same property under different identities to manipulate market perception.
Visual Understanding
Beyond matching, Amenta understands what it sees. Room type detection: living room, bedroom, kitchen, bathroom, balcony, hallway. Amenity detection: pool, gym, parking, garden, elevator, security. Property type inference: apartment, villa, duplex, studio, office, retail.
This visual understanding supplements the textual listing data. When a listing claims 3 bedrooms but the images show 2, Amenta flags the discrepancy. When a listing omits the parking but the images show a garage, Amenta adds it.
Visual Change Detection
A property listed twice — six months apart — with updated photos. Amenta compares the image sets and detects: renovation completed (new flooring, new paint, new kitchen), furnishing added or removed, or condition deteriorated. This temporal visual analysis provides context for price changes that text alone cannot.
Condition Assessment
From visual indicators alone — image quality, finishing level, fixture conditions, visible wear — Amenta generates a visual condition score. This score supplements the price and location data with an additional assessment dimension, enabling more accurate comparable analysis and valuation.
Property Search and Monitoring
Search the entire market through a single interface: by district, price range, area, property type, rooms, or visual attributes. Save searches and receive alerts when matching properties are listed, when tracked properties change price, or when market signals fire for monitored districts.
On-Premise Architecture
Amenta runs on your infrastructure. Property data, images, analytics, and alerts remain inside your network. No cloud dependency for any function. For banks and government housing departments where property data is sensitive, this is the only acceptable architecture.
Part of the ARMANET Platform
Amenta shares the same AI vision foundation as ARVORA (vehicle intelligence) and NAREX (security intelligence) — the same vision models, adapted for real estate imagery. Through the Foblax research partnership, Arabic-language queries and Arabic reporting are native capabilities. Amenta is the market intelligence layer of the ARMANET ecosystem.
Who It Is For
- Real estate developers: Price your launches based on real market data. Track competitor inventory. Detect demand shifts before they become obvious.
- Investment firms and REITs: Identify undervalued assets. Assess risk before acquiring. Monitor portfolio performance against market benchmarks.
- Banks and mortgage lenders: Automate collateral valuation with comparable evidence. Detect inflated listing prices. Monitor market risk exposure.
Government housing departments:
- Monitor housing supply, track affordability trends, and detect market manipulation across districts.
- Brokerage firms: Manage inventory accurately. Detect relistings and stale properties. Provide clients with evidence-based pricing.
The Competitive Position
No platform in the Egyptian or regional market combines property data aggregation, deduplication, price intelligence, and AI vision in a single system. International platforms like Zillow or PropertyGuru are consumer listing sites — not intelligence tools. Regional portals show listings; Amenta analyzes them.
The AI vision component — image matching, deduplication, visual understanding — is unique to Amenta. No competitor in this market offers it. And no competitor presents the analysis in Arabic.
