Data Origin & Methodology

1. Purpose of This Document

Transparency about data sources, transformations, and methodology is critical in real estate investment analysis. This page explains how Estara AI collects, aggregates, and processes data to power our analysis, forecasts, and tools, and clarifies important limitations and assumptions.

2. High-Level Data Architecture

Estara AI combines information from multiple categories of data sources:

• Public records and government datasets (e.g., property tax, assessor, and geographic data)
• Licensed third-party data providers (e.g., HasData, SchoolDigger and other property, market, and school data vendors)
• User-provided inputs (e.g., purchase price, rent assumptions, operating expenses, financing terms)
• Derived metrics and AI-generated analysis created by Estara AI's proprietary models

These sources are combined to produce investment metrics, forecasts, risk indicators, and narrative insights.

3. Public and Government Data

Many core property attributes are based on public records, including:

• Parcel identifiers and property location
• Land and building area (square footage)
• Property type and use codes
• Assessed values and property tax data
• Recorded sale dates and sale prices (where available)

Public data is typically obtained from local governments, county assessors, recorders, and open data portals. Such data may be incomplete, delayed, or inconsistent across jurisdictions, and may need normalization or cleaning.

4. Licensed Third-Party Data Providers

Estara AI also uses licensed third-party data providers to enhance and enrich property-level and market-level insights. These providers may include:

• HasData: Property-level details, listing metadata, and market attributes where permitted by license.
• SchoolDigger: School locations, catchment relationships, and school performance metrics.
• Other analytics, mapping, or demographic data vendors as necessary.

Each provider has its own methodology and limitations. Estara AI does not independently audit every data point and relies on providers' documentation and update processes.

5. User-Provided Inputs

Many investment outputs depend on assumptions and data that you supply, including:

• Acquisition price and closing costs
• Down payment and financing structure
• Interest rates, amortization periods, and loan terms
• Expected rents, vacancy assumptions, and rent growth
• Operating expenses, maintenance, and capital expenditure plans
• Exit cap rates, holding periods, and target return thresholds

Because these inputs are under your control, Estara AI cannot and does not guarantee that any particular scenario or projection will match actual future performance.

6. AI Models and Derived Metrics

Estara AI uses AI and algorithmic models to generate derived metrics and insights, such as:

• Cash flow projections and multi-year holding period analysis
• Net operating income (NOI) and capitalization rate estimates
• Internal rate of return (IRR), cash-on-cash, and equity multiple calculations
• Risk indicators (e.g., sensitivity to vacancy, interest rates, or rent changes)
• Market trend narratives and qualitative summaries

These outputs are generated based on a mix of user inputs, third-party data, and model assumptions. They may not fully capture all property-specific risks or local market nuances.

7. External Listing Links and Non-MLS Data

For user convenience, Estara AI may provide links to property listings on third-party real estate portals such as Zillow, Redfin, Realtor.com, Homes.com, Trulia, or similar sites. These links are provided as references only.

Estara AI does not directly access, scrape, ingest, store, or distribute data from Multiple Listing Service (MLS) systems or any MLS-protected databases. Any information that appears similar to listing data is either:

• Derived from licensed third-party providers
• Based on public records and open data
• Entered by users themselves

Estara AI is not affiliated with, endorsed by, or sponsored by Zillow, Redfin, Realtor.com, Homes.com, Trulia, or any other listing platform and does not control their content.

8. Geographic Coverage and Scope

Data availability varies significantly by region, state, and county. Some metrics or visualizations may not be available in all areas due to limitations in public records, third-party coverage, or data quality.

Where coverage is incomplete, Estara AI may display partial data, estimates, or omit certain visualizations. Users should interpret such gaps cautiously and, where appropriate, supplement with local knowledge and professional advice.

9. Data Refresh and Update Cadence

Different data sources update at different frequencies. For example:

• Tax and assessor data may be updated annually or periodically
• Market and rental data may be updated monthly or quarterly, depending on provider
• School performance data is often updated annually

Estara AI strives to incorporate updates from providers as they become available but does not guarantee real-time data. There may be delays between a change in underlying data and its reflection within the Service.

10. Assumptions and Model Limitations

All models rely on assumptions. Examples include assumed holding periods, rent growth rates, expense growth, vacancy rates, and financing terms. While you can customize many of these assumptions, others may be embedded in our models and not visible at the surface.

Model outputs are inherently uncertain and may diverge materially from actual outcomes due to:

• Unexpected property-specific issues or capital expenses
• Macro-economic changes (e.g., interest rates, inflation)
• Regulatory changes (e.g., rent control, zoning, tax changes)
• Data errors, missing data, or stale information

11. Interpretation of Estimates and Scores

Estara AI may present estimates (such as values, yields, or risk scores) using ranges, scenarios, or single-point estimates. These outputs should be interpreted as directional indicators rather than precise predictions.

Users should not rely solely on any single metric, score, or forecast when making investment decisions and should instead consider multiple scenarios, sensitivities, and professional guidance.

12. Reconciliation with Other Data Sources

You may observe differences between Estara AI's numbers and other sources (e.g., MLS reports, broker pro formas, other platforms, lender underwrites). These differences can arise from:

• Different input assumptions and time horizons
• Use of alternative data sources or update frequencies
• Rounding conventions and calculation methods
• Data quality issues or missing inputs

Where discrepancies are material, users should review underlying assumptions and consult with advisors to reconcile the differences.

13. User Responsibility and Independent Verification

Estara AI provides tools to support analysis but does not replace due diligence. You are responsible for verifying critical details, including:

• Actual rents, lease terms, and rent rolls
• Operating expenses, property condition, and CapEx needs
• Local regulations, zoning, and compliance matters
• Financing terms, lender requirements, and closing costs

You should independently verify data with local professionals, including real estate agents, property managers, attorneys, accountants, and lenders before making investment decisions.

14. No Guarantee of Accuracy or Performance

While Estara AI aims to provide useful and reasonably accurate tools, all data and analysis are provided "as-is" and "as-available," without any warranties or guarantees of completeness, accuracy, or suitability for any particular purpose.

Nothing in this document or within the Service should be construed as a guarantee of investment performance or as a promise that any particular strategy will succeed.

15. Changes to Data Sources and Methodologies

Estara AI may modify its data sources, models, assumptions, and methodologies over time as we improve the Service, onboard new providers, or respond to data availability changes. These changes may cause historical results to differ from current outputs for the same property or scenario.

We may update this Data Origin & Methodology page periodically to reflect significant changes; however, we are not obligated to describe every minor update or internal model adjustment.

16. Relationship to Terms of Service

This Data Origin & Methodology page is intended to provide transparency and context and does not modify the Estara AI Terms of Service. In the event of any conflict between this page and the Terms of Service, the Terms of Service shall control.

17. Contact and Questions

If you have questions about our data sources, methodologies, or assumptions, please contact us:

• Email: legal@estara-ai.com