AI Solutions in PropTech: Overview
Published on
Overview
AI Solutions in PropTech by Smart Indexes enable location search and property selection through a CMA (Comparative Market Analysis) workflow tailored for Dubai. The approach combines Cap Rates by communities, master projects of development and buildings with risk layers such as Flood Risk (FR), Average Crime Rate (ACR), and Traffic Accident Rate (TAR). For asset-level precision, property characteristics and contextual analytics are retrieved via the Makani Number.
The result is an explainable, data-driven shortlist of candidate assets and locations, aligned with investor or developer objectives (yield, stability, payback horizon, compliance constraints).
What the CMA workflow includes
Cap Rates: retrieval and comparison by communities, master projects and buildings.
Makani-based property profile: land use / building usage, gross area, height, immediate amenities, points of interest, social and transport accessibility.
Risk layers: Flood Risk, Average Crime Rate, and Traffic Accident Rate integrated into a unified scoring model.
Explainability: transparent ranking with metric by metric justification for each item in the shortlist.
Data inputs & Makani Number
Makani is used as a stable geo-identifier for buildings and plots in Dubai, enabling precise joins between core property attributes and spatial layers. Smart Indexes exposes Makani-based endpoints to collect characteristics and environmental context required for consistent CMA.
| Input | Description |
|---|---|
| Cap Rate | Rate values by community / master project / building used for yield comparison. |
| Makani-based attributes | Size and usage, building height, proximity to amenities, transport accessibility, social infrastructure. |
| Flood Risk | Geo-referenced probability of flooding for the given location. |
| Average Crime Rate | Aggregated crime intensity for the surrounding area. |
| Traffic Accident Rate | Incident density around the asset or within administrative boundaries over a defined period. |
Methodology (high level)
The AI agent orchestrates data via Smart Indexes API and produces a composite score for each candidate location or asset:
Scope selection: user specifies the intent (e.g., «office with high yield and lower risk within 15–20 minutes of key business hubs»).
Data retrieval: the agent queries Cap Rates by geography and gathers Makani-based property details, then overlays Flood Risk, Average Crime Rate and TAR.
Normalization & weighting: metrics are normalized to comparable scales; weights reflect the strategy (yield-first, safety-first, balanced).
Ranking & explainability: assets are ranked; each result includes factor-level contributions and threshold checks (e.g., max acceptable risk levels).
Why it matters
Higher confidence: combines yield and environmental risks at the specific asset location.
Operational speed: automated shortlisting reduces manual pre-screening and analyst load.
Consistency: Makani-centric joins ensure stable and repeatable asset evaluation.
Better trade-offs: identify assets with attractive Cap Rates and acceptable combined risk.
Typical use cases
Investors: shortlist buildings in communities with target cap rates and low combined risk within 1 km around the asset.
Brokers: generate explainable proposals with standardized metrics and map-based evidence.
Developers: compare plots within master developments by yield potential and exposure to environmental and safety risks.
AI/IT teams: power chatbots and agentic workflows that answer «why this asset» with metric-level transparency.
How it looks in GeoAnalytics
All layers are available for Dubai in Property > Summary > Risks and related sections. Users can toggle Cap Rates, inspect Makani-based details for a specific building, and overlay Flood, Crime, and TAR to refine decisions.
Notes
The overview focuses on analytical metrics and orchestration. It does not rely on transactional «deal history» data; such records belong to other systems (e.g., Dubai's EJARI API) and are outside this workflow.
Makani-based joins are essential for consistent asset-level analytics and should be used whenever available.