houzen made quite the impact when it hit the property sector back in 2017 with its vision to become a global real estate exchange. Using AI to match properties with the right tenants and buyers, all in five days or less, proved to be an effective solution. houzen helped institutions to rapidly lease and sell properties two to three times faster than the rest of the market, as well as producing 10-18% higher rental and sales offers than the local market standard.
Now, all eyes are on the company’s new AI valuation tool. Using AI true value analysis, the new AI valuation tool analyses the entire UK’s property stock in seconds, using customised buyer criteria to match investors with the top 1% of properties that most suit them.
Using automated deal picking and factoring in gross and net yields, returns (IRR, multiples), calculations are led by computer vision and proprietary data on live closed sales and rent deals. The results are then examined by a team of expert human analysts to add further value and reassurance in the findings. Each property is pressure-tested on several investor metrics, so that users can make faster go/no-go investment decisions.
The impact for investors is certainly noteworthy. Their properties are matched with highly focused ‘warm’ investors, cutting out wasted time and delivering a slicker, faster investment process. houzen’s data-rich platform accounts for a wide range of investment strategies, such as sustainable homes, flipping and high yields. It’s win-win with artificial intelligence, big data and computer vision leveraged to deliver high quality matches.
“We’ve combined accurate and exhaustive real-time data searches with a machine learning and computer vision methodology and human oversight that means the process gets better with each iteration. With every step,, our emphasis has been on delivering benefits to all those involved in the investment process – not just the buyers but the vendors too. Each party can benefit from smoother, faster transactions as a result of using our new AI valuation tool.”
houzen’s new AI valuation tool sources its constantly refreshed data from the UK’s real estate listings portals (such as Zoopla), along with independent property data aggregators such as PropertyData, healthy street index, tranquil city and others focused on biodiversity, noise, pollution and neighbourhood data. It also draws on Land Registry data for closed prices and proprietary data on closed sales and rents through houzen’s 3,000+ agent and developer partners. That proprietary data is the company’s secret weapon when it comes to price prediction analysis.
Automated machine learning and computer vision models sift through a host of digital information including pictures, floorplans, text, videos and more. Manual quality checks by real estate and private equity investment professionals support the ongoing iteration of houzen’s algorithms. This process drives up the quality of the data and the modelling, removing any biases that the machine may have developed. Over the eight months that houzen has been building and enhancing the programme, the model’s accuracy has increased from ~85% to close to 95% accuracy in over 800 UK postcodes.
Rental models, meanwhile, are currently running at equally impressive 89-92% accuracy levels. Work continues, with houzen’s data science team shooting for an industry-leading error rate of <2%. Accuracy of this level goes a long way to providing everyone from lenders, developers and architects to investors with the confidence needed to make key decisions in their respective area of work.
“We’re already at an exciting stage with our new AI valuation tool and, with our machine learning and computer vision models in place, the results are continually improving in terms of their accuracy. The world is changing rapidly around us. The houzen AI valuation tool enables all those involved in the investment process to respond to those changes rapidly and in a data-driven way.”
Privacy & Cookies Policy
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.