Executive Summary
Property buyers are deciding on long-term commitments, and the credibility of the investment is crucial for banks that provide mortgages. Every property purchased with the bank's involvement must be inspected by an expert, who verifies its condition and value with the results presented in the form of standardized documentation.
In order to deliver these quickly and efficiently to the bank, Value AG decided to digitize the process in-house by building a valuation support system called Easy Valuation Application, or EVA for short. We were entrusted with the development and creation of the text and research module.
About Value AG
Germany's largest real estate valuation company, Value AG, has been in business for over 20 years. They develop technology to reshape The valuation of the real estate market place. Working with IT experts allows them to continually improve products to create a digital, efficient, and flexible valuation process.
Intentions
- Digitizing the process by building a valuation support system - EVA
Outcome
- A solution dedicated to product owners for the future development of the tool
- Reduce the time needed to create the evaluation
- Create one place to search for data instead of multiple sources, pages, and files
- Possibility to check which features are used by valuers
- A tool presenting data similarly to Google Analytics
- The module shows user behavior, the number of visits to the tool, and the user's location on clear graphs
EVA Research module
The Application is a comprehensive knowledge base for valuers. It is built on data that has been categorized to give a complete picture of the property and its surroundings so that an efficient valuation can be achieved.
The data is presented in such a way that it is clear, consistent and can be useful when using easy-to-read maps, charts or even simply by generating full content that can directly be used in the final assessment.
How does it work
The EVA Research module is a web application that presents location data based on the address searched by the user. The application shows data on different types of properties, sources of noise and emissions, public transport, roads, infrastructure, prices of comparable properties, and much more.
Data necessary for valuation in the one application:
- Maps with information for valuers: showing, e.g., noise pollution levels in neighborhoods, quality of life conditions
- Neighborhood infrastructure: POIs, population, information about restaurants, pharmacies, schools, prisons, etc.
- Useful statistics: information about certain spots, GDP per person
- Description of the real estate market: how it looks like, how it has changed in recent years, what the prices were (renting, selling)
- Valuation database
- The indication of real estate values and a value of capitalized profits
- The indication of a comparative value
- Location description
Automated generation of maps
We've received other challenges with Research: we do minor services that are used to build a full property valuation:
- MapSets - PDF and Word files
- Non-official cadastral maps
The valuers can generate maps automatically based on the available data, which in turn reduces the time needed to complete the work. Previously it took about 10 minutes to prepare the document with the necessary detailed information. Now it's just a simple click away.
Key values
- Reduce the time needed to create the evaluation
- One place to search for data instead of multiple sources, pages, and files
- Minimize errors
- Maintain up-to-date data
The challenge
In order to successfully develop the application, it was important to fully understand the industry and the activities of the experts preparing the valuations.
Our team took on an important technological challenge. We chose to run in Azure so that we could manage the technology freely, even without full knowledge of how the client's current system works. We used Azure Functions, an on-demand service, to create a fully updated infrastructure and resources needed to run the application.
The priority for us is to match the technology to the needs, function of the tool and goals. When we created the maps, we chose different technologies depending on their specific features.
We faced
- Data quality - deduplication, e.g. if a user asks about the two closest bus stops, we cannot give him the same stop but in a different direction
- Data normalization and interpretation - all data is searched by one address, the system must understand all data in the context of location
- The efficiency of data searching - use of caching, immediate response
- Data mining - various formats, multiple APIs, excel sheets, data packages.
- Automatic updating of data periodically
- Working with unstable data sources - retrievals, monitoring, graceful error handling
- Presentation - complex data in a simple form
Solution
Value AG wanted to know the needs of its application users. We created a solution dedicated to product owners for the future development of the tool. Thanks to the data, they can check which features are used by valuers, Thus knowing which parts of the system are more important to them as a result. We based it on Azure Application Insight, which means that we created a tool presenting data similarly to Google Analytics. The module shows user behavior, the number of visits to the tool, and the user's location on clear graphs.
Result
As our teams have been involved in working on some parts of EVA, we feel that our work together has an impact on the development of the modern real estate market. The application has a real impact on the quality of valuers' work and increases their efficiency. Automation is game-changing, even though initially it may have raised fears that the technology itself would be taking their jobs away they realised it in fact made their jobs easier.