Selling real properties to demanding customers using Google Cloud Platform
Prontopiso is the only real estate agency that puts money in your account before the sale if you need it; gives you the time you need to find the best buyer; and is by your side from the moment your house is valued until the moment you sign it over before the notary.
Ever since Prontopiso was first imagined, it was crystal clear it aspired to be more than just a showcase for beautiful properties everyone wanted to live in.
Industries: Real Estate
20 September 2019
Hi, Gonzalo, thanks very much for your time and for choosing us to carry out your GCP project. We’d like to start by finding out a bit out the challenges ProntoPiso faced before you started working with Kubernetes.
Hello! I’m delighted to be able to team up with you guys in promoting the benefits of working with Google Cloud Platform services. ProntoPiso is a real estate agency that is looking to revolutionise the traditional property sector. We worked out that one of the most successful strategies for us right now was to open up different areas for agents in Spain from a geographical standpoint.
Take a look at Zaragoza at the moment, for instance. We don’t have a physical ProntoPiso premises there but we want to give all the independent agents an individual stack which allows them to bring their properties on to our systems. From there, ProntoPiso can value the property to assess whether or not we can offer them a guarantee. Ultimately, the guarantee would mean us paying 100% of the price of the property to the customer ahead of time.
We saw a need and a challenge that we were unable to solve easily with the platform we had before.
Lack of scalability by location
Having noted this, we got in touch with Nubalia as Google Partner and we saw that Kubernetes would give us the option of creating multiple environments within a single server in order to cater for these smaller areas. It’s a case of adding a subdomain like Zaragoza.piso.com where agents can then work in a completely separate environment to manage only the properties in that city. That was the challenge – being able to scale all the locations across the housing stock available in Spain in record time.
It 100% sounds like a challenge for the Cloud. Do you think you’ve achieved it? I mean, are you satisfied with the scalable potential that Kubernetes provides right now?
Just today, we put the platform into operation on GCP with the first stack in Valencia and the results so far have been excellent. For testing, we’re opening stacks in different cities and the plan now is to open in Malaga, Bilbao and Seville. For us, it’s as simple as clicking a button, opening the location and the area will automatically become available with the Kubernetes environment set up easily in 5 minutes. The business team are even happier than us in that regard (laughs).
Possibilities – Stand out from the crowd
Well, ultimately Google Cloud Platform has something to bring to every department. In your case, what have the benefits been of working through a partner and not directly with Google?
In our case, Enrique (Cloud Director Nubalia) and Yolanda (Apigee Lead Google) got in touch for the first time and they came down here joined by Google, which makes quite the impression. We used to be a hub here with everyone working with Amazon Web Services. The Google brand really makes you sit up and listen to what they have to offer. When you start looking through the full selection of services offered by Google Cloud, it’s easy to get overwhelmed with the vast array of things you can sign up for. To be guided by a partner in this case, with all the advice they provided at every stage, made everything much easier (not to mention all of the specific knowledge of the Google services they were offering aimed at precisely what we were looking for on a business level). I think the most difficult thing when you turn up at Google’s front door, is exactly that, knowing which door to open first; such is the quantity of services and tools available.
That’s right, sometimes it can be a real mission if you take the plunge into this world without the right guidance. You mentioned you used to work with another cloud, so what inspired the change?
Right, the explanation for that is extremely straightforward, it’s what makes Google stand out from the crowd, it’s just what I was saying about ProntoPiso. We pay customers 100% of the value of their property in advance. We take a fairly big risk in doing so because if we haven’t sold the property in 90 days, we’ll automatically pay out the value using ProntoPiso funds.
So on the cloud, we weren’t only looking for Kubernetes but we also needed the data side. It was the Google solution integrated with Kubernetes which provided all the data services in such a simple fashion that finally made us say: “Right, let’s make the jump, we’ll go with Google”.
So far, we really like what we’re seeing, we knew it wouldn’t be a case of only Kubernetes. Aside from that, what’s really pleased the data scientists we’ve got here is Big Query. They’ve really fallen in love with the ability to build models connected directly to the services we’ve got in Kubernetes. Essentially, that’s the biggest benefit of Google Cloud, just how easy it is to integrate data and structure.
And now that you’re all in, are you planning to start any other GCP projects?
Since day one with Enrique we were keen on a project he was already working on in Mexico. We’re keeping an eye on Google’s Vision API because that’s one of the missions we’d like to embark on at ProntoPiso. I mean, everyone knows that when you’re at home and someone knocks the door and you see it’s a real estate agent, your first thought is, “Not again!” People don’t like it, ultimately it’s an overcrowded sector. What we want is to make this part, the property valuations, a bit more fun and like real life.
Now we’ve trying to work on a project related with Vision API. We want to get to a point where a customer is valuing the property and they can upload images which would serve to help formulate the price of the property. For instance, if they could assess the amount of light in a flat. One of the issues we have is that we’ve got macro information on the portals, we’ve got market information on properties, but nobody knows the condition of the inside of a property, nobody has been able to get inside and look around. We want to be able to get inside those properties to ultimately verify the types of characteristics they possess, which then, using statistics, could be extrapolated to surrounding properties fairly easily using some interesting models. The other side of this is you have to make a tool that is almost like a toy to the customers, so they can see how they upload pictures of their properties that automatically have an influence on the valuation.
Ultimately, this is one of the projects that has got us most excited. We’re going to put a load of energy into this, we want to prepare lots of different models, to the degree we can say: “Let’s enter all the images of ovens that exist on Amazon”, for instance, all the stovetops, extractor fans, sofas, lamps, to ultimately determine the quality of these qualitative variables contained inside a property which we are unable to see unless our agents go and visit it. That’s what we want to be able to transform and digitise as data in the end.
This tool would allow us to enhance all kinds of things, so we are continuing to work hard and we’re still establishing the best way and exactly what it is we want to do but we believe it could happen fairly simply and in the short term thanks precisely to Vision API. It’s not just about the ready-made tool Google provides anymore, but also how you can work on and develop what you want, which is the most amazing thing for a business model.
Wow, that’s some incredible stuff, Gonzalo. You can really sense your excitement for the project and it’s our pleasure to have the chance to be involved with developing something like this. I can’t even imagine what’s to come next!
Yes, the plan is to keep making progress and doing new things because Google is always taking enormous steps forward. And we’ll task Enrique with helping us complete the preliminary analysis and giving us a few clues, but that’s for the future.
Thank you so much for sharing ProntoPiso’s experience with us, Gonzalo.
No problem at all. Thank you for giving us this opportunity and for being by our side, helping us progress with this project. Ultimately, the result will be thanks to everyone’s efforts, that’s for sure.