Online Home Loans is implementing AI to guide customers to the best mortgage

- March 29, 2017 3 MIN READ
Online Home Loans

With 2017 seeing artificial intelligence (AI) emerging to help you decide between cheesy crust and plain crust on your pizza, it seems that at least some of the predictions for AI this year are coming true.

Although no one directly foresaw the uprising of Domino’s DRU Assist, voices in the tech space did predict an acceleration of ‘chatbots’; AI that can interpret human languages, either written or spoken, and guide a user to or through a product or service.

Online Home Loans is one such startup, looking to harness AI in order to make a splash in the, well, online home loans industry.

With the platform beginning its run as a traditional home loan comparison site similar to other Aussie sites ClickLoans and Lendi, the startup’s founders Murray Lees, Cameron Upshall and Leanne Brinkies, saw both within their own model and others in the market that there was no simple way to “easily” apply for home loans online.

“You can find a lot of information and comparison offerings, but easy way to actually purchase a home loan after researching it. We make it easy for a person to do the research and keep on clicking at the end to apply,” said Lees.

Lees’ experience in the mortgage broking industry stretches back decades, having worked in finance for roughly 33 years; he registered the platform’s domain name during the dial-up age of 1999.

Lees said his time in the industry taught him a lot about what customers are looking for and how to scale their experience, experience that would eventually lead to the development of a mortgage-savvy AI named Sophie.

“Our whole tech proposition is to take a new user exploring this unknown space and transition them to becoming a customer, understanding things and connecting with a vendor,” explained Cameron.

Planned to launch alongside the new model for the site in the coming weeks, Sophie will be present as a chatbot available for users to guide them towards a home loan they’re qualified for based on their responses to a series of questions.

Currently, the site offers a human assistant for users to chat to in addition to a help line number. According to Cameron, rather than replace the human assistance entirely with AI, users will still be able to converse with someone if Sophie detects things are “complex”.

“She can them guide them to a home loan or detect if the user has a more complex scenario, by which she’ll guide them to human support,” he said.

“Everyone can relate to not wanting to call a company and a lot of people want to be able to do everything online, so Sophie’s there to offer a simple alternative. She doesn’t replace a human, but really augments that experience.”

Built on IBM Watson, Cameron explained that Online Home Loans has also developed its own machine learning system for Sophie to learn off.

Learning will function by the startup’s team ‘feeding’ Sophie user data based off their searches and questions every six days, in order to train the AI’s responses.

The goal, according to Cameron, is to get Sophie to a point where she can guide customers through the entire online process, beyond just showing what they’re qualified for.

Beyond introducing Sophie, the site’s new model will also allow customers to apply for home loans from over 30 lenders once they’ve sorted through the comparison process.

To verify a loan, Cameron explained that Online Home Loans has put together its own verification process which takes all the required information and documents from the customer before lodging them to the appropriate vendor. Similar to a traditional broker, Online Home Loans then gets a cut of each successful loan from the residing bank.

With a slew of new features set for the new model, Murray said the platform will still retain nearly all of the features on the existing platform, including a borrowing power, repayments and stamp duty calculators, amongst others.

Every feature counts as the competition in the space only continues to grow; fellow startups like HashChing, which has helped over 10,000 borrowers, are coming at the home loan space from fresh angles to help find consumers a better deal.

To help increase Sophie’s “sophistication” and fund development of the platform, the startup is currently in the middle of a Series A round, with Cameron saying the business is looking to raise $1.75 million.

Image: Murray Lees & Cameron Upshall & Leanne Brinkies. Source: Supplied.