Have you ever taken a different route home because Google Maps warned you of heavy traffic in your area? Or your car’s tracking device automatically alerted your insurer that you’ve been in an accident because it noticed unusual driving behaviour.
Machine learning is no longer just a buzzword in the world of technology; its applications and uses have started to affect our everyday lives without us even noticing.
Using Netflix’s algorithm to pick a movie to watch, using Uber or asking Siri what the weather will be like tomorrow all make use of machine learning.
“It’s already here and all around us,” says Zabeth Venter, CEO and co-founder of Averly, a platform that uses machine learning to simplify and improve the relationship between tenants, landlords and rental agents alike.
Machine learning is an area of computer science that involves developing and deploying algorithms to provide a software program with the ability to learn without being explicitly programmed.
In Averly’s case, machine learning is used to extrapolate a prospective user’s convictions and opinions about renting by having them answer a bespoke survey that will not take more than a few minutes to complete.
As more people complete the survey, the algorithm becomes better at interpreting the results and more accurately predicting the specific user’s expected behaviour.
This is merely one example of machine learning being used for good, says Venter. Averly uses machine learning to help unravel the reasons why a landlord and his or her tenant often have a strenuous relationship.
“If you know the consequences of kicking a hole in a door won’t be good, you’d stop doing it.” However, tenants often do not realise what those “kicking a hole in a door” actions are that can strain their relationship with their landlord.
The same logic applies to landlords and agents acting in incongruous ways that are upsetting tenants.
“It can be something as simple as a landlord not paying the utility bill and the tenant having no access to hot water. It’s oftentimes small things that you would think are obvious, but in reality, often falls to the wayside.”
Averly does not make decisions on behalf of landlords, agents or tenants. The platform merely acts as an unbiased informer based on the information it receives from each person that fills out the survey.
“Averly collects your data. You are not just one of, say 200 million. It interprets the results based on your context,” explains Venter.
It also never takes sides. “It will only highlight potential aspects of how you think about, for example, renting a property and how that can influence your experience during the rental process.”
Having this knowledge before a rental contract officially starts can help a landlord, rental agent and a tenant to pre-emptively address any concerns ensuring a better overall experience.
During the survey, a prospective tenant can, for example, indicate that he or she does not think paying their rent on time is such a big deal. By having this knowledge, a landlord can make arrangements with the new tenant on what steps will be taken should they ever pay their rent late.
Knowing possible risky future behaviours can help to avoid putting unnecessary stress on the relationship between tenants, agents and landlords.
“The fundamental building block of using machine learning in this way is to create trust.”
What about my data?
Averly only collects data that a prospective tenant would have provided to a landlord or rental agent anyways, says Venter.
This includes financial information like bank statements. However, these documents are not analysed by Averly’s machine learning algorithm and are merely stored in an online repository for easy access by landlords and agents.
Some verification software is used to make sure the statements provided are true and accurate.
“We contain all the information in the Averly bubble and will never sell it to anyone. We will only prompt you to be better at your role in the rental space. So, if you’re an agent, to be a better agent; if you’re a tenant to be a better tenant and the same for landlords.”