What is a Machine Learning Automated Valuation Model?

Valeri uses proprietary machine learning systems to analyse a massive array of data points, producing an Automated Valuation Model (AVM) with industry leading accuracy in real time.

Valeri’s accuracy is backed by years of work by our data scientists, perfecting the right combination of data layers, artificial intelligence technology and proprietary algorithm development.

Valeri Automated Valuation Model

image/svg+xml

Market First AI Driven AVM

Valeri is the market’s first Artificial Intelligence powered automated valuation model (AVM). Valeri provides more accurate results in less time, resulting in improved risk management, and Loan to Value Ratio forecasting.

Legacy Automated Valuation Models

image/svg+xml

Old Technology & AVM Methodology

Legacy AVM's are driven by historical sales and local indexation. The use of fixed and manufactured metrics results in overstated confidence.

Valeri Automated Valuation Model

image/svg+xml

Superior Accuracy

Valeri’s accuracy and secure processing systems are consistently validated through weekend blind testing and back casting models.

Legacy Automated Valuation Models

image/svg+xml

Accuracy has been Compromised

Traditionally, the primary objective of AVM's has been to sell loans and reports. The heavy reliance on historic sales data has produced inaccurate valuations.

Valeri Automated Valuation Model

image/svg+xml

Innovative Metrics

Our innovative Land Value Algorithm accounts for geographic & socioeconomic features down to individual land parcel level.

Legacy Automated Valuation Models

image/svg+xml

Limited AVM Metrics

AVM's rely on arbitrary area boundaries and wide-scale trends. Land values are not independently considered.

Frequently Asked Questions

Yes. As advanced as our machine learning and Artificial Intelligence systems are, they can pick up poor habits which affect accuracy. The accuracy of a machine learning powered AVM depends on the quality of the data it receives and the processes of the machine learning system itself. Think of it like maintaining a race car – if it is given high quality consumables and is finely tuned by experts, it will perform. If it’s given cheap consumables and is neglected, it’s performance will suffer.

Once a machine learning AVM is perfected, we systematically train it on specific cities with clean, city-specific datasets. This allows Valeri to learn the patterns, values, and market aspects unique to every city. Because Valeri is always learning and always being trained, it will only get smarter.

PointData has developed its own proprietary training data layers that are contiguous to specific locations, from neighbourhoods to entire cities and regions. Under strict supervision, we train Valeri with location specific data in order for it to identify and learn the patterns and insights unique to each city. These proprietary data layers do not form part of a sales dataset, yet they are just as important in improving Valeri’s accuracy.

PointData is constantly researching new and evolving areas of the exciting world of Artificial Intelligence to ensure the accuracy of the algorithms behind Valeri.