Snowflake makes the perfect home for CoreLogic's real estate data

There are few industries as vast as the real estate industry. There are nearly 200 million parcels of land in the US and about 1.5 million real estate agents ready to buy or sell them. Real estate is the single biggest asset class in the world, and bankers and insurers are carefully, constantly tracking its value.

That value is influenced by a huge number of factors. The value of each parcel depends on variables like the changing climate, the local economy, the structures on the land and more. CoreLogic is a data analytics and technology company that underpins the real estate industry, helping realtors, insurers and lenders understand the market. After more than 10 years in business, CoreLogic is bringing its tools to Snowflake — because that's where its customers are. CoreLogic Chief Innovation Officer John Rogers explains it this way:

Obviously, getting access to data, our models and all things property is of great importance to our clients so that they can grow their business and mitigate risk. Part of our strategy is to provide choice and flexibility to our clients. And what that means is really meeting them in their world, in their cloud stack and their tech stack. One of those important worlds is Snowflake. It's very quick to spin up, leverage data assets and perform analysis for the company that uses it.

CoreLogic moved all of its own infrastructure to the Google Cloud Platform, in a process that began about three years ago. Since then, Rogers says, the company has seen massive uplift in terms of scale, processing speed and security. And of course, it now has access to GCP's AI and ML capabilities. With GCP underpinning its own operations, Rogers says, CoreLogic understands what its own clients are looking for from cloud infrastructure and platform providers.

In sectors like financial services, platforms like Snowflake and Databricks offer the multi-cloud operational efficiency needed. However, as highly-regulated sectors adopt multi-cloud strategies, it becomes all the more imperative for CoreLogic to meet its customers in their own environments. Rogers says:

Transferring data around from behind the clients' four walls is time consuming and obviously has security and compliance challenges. A lot of those disappear because now we're working in their world, within their Snowflake instances. So that takes away a lot of the pain for our financial clients, which is massive.

Bringing property data into Snowflake

To get to Snowflake, CoreLogic turned to Hakkoda, a specialized systems integrator that could reliably and quickly build out data models and applications within Snowflake. The whole process took just about three months and was completed without a hitch. Rogers says his team would have tolerated some friction in the process, but it wasn't worth the effort for CoreLogic's own teams to go at it themselves. He says:

We would obviously love to learn as a company, but I don't want to learn every single mistake as we forge a new path into a new technology stack like Snowflake. Hakkoda really provided the expertise, the know-how and the drive to set us up in Snowflake... I think doing it ourselves would have taken double the time. We have very talented people, but there's always a learning curve. Speed to market, as well as supporting CoreLogic data engineers and data scientists, was key for me, and working with a specialized SI like Hakkoda just fits the bill perfectly.

The first CoreLogic app that Hakkoda built natively for Snowflake is called Clip (CoreLogic Integrated Property). It's a program that gives every US property a unique identifier, linking all instances of that property across data sets. It's effectively like a Social Security number for properties. Users can get a comprehensive understanding of any property, even ones that have yet to enter tax-roll records. It shows historical data like border adjustments and property splits, property structure changes, ownership changes and tax changes; it includes valuations, environmental data, geospatial data, and other variables. Users can easily 'clip' data from a set of properties and use it to generate unique insights within their Snowflake instances.

CoreLogic already has dozens of clients using Clip to access data via Snowflake. Rogers says:

All that data integration really allows clients to focus on the insights rather than the data wrangling. Property identity is a hard problem to solve, and we've spent many years [solving it]... We're now monitoring the speed of setup, the speed at which they can clip their portfolio, and analyze data. Because we want to do it in minutes and hours rather than days, weeks, and unfortunately sometimes months, historically. That's what we're measuring actively every day.

Bringing Clip to Snowflake is just the start. CoreLogic plans to continue working with Hakkoda to roll out other capabilities on the platform, such as property climate risk analytics. With this functionality, customers can click on any property in their portfolio and understand the risk and financial exposure that the changing climate will create by the year 2050.

CoreLogic also plans to build out its precision marketing tools, allowing lenders and other companies to target individuals with the right message at the right time. By combining property data with other sources of information like mobile phone usage, CoreLogic customers will be able to get a sophisticated, fine-grained understanding of their potential customers' behavior patterns.

My take

Now that multi-cloud operations have become commonplace, it's imperative for vendors like CoreLogic to meet their customers where they are. There's sure to be a growing expectation that tools can work across platforms, without running into regulatory hurdles or creating security risks.

The challenge comes when vendors have to task members of their team with rebuilding applications, taking up their valuable time and making them vulnerable to the pitfalls that inevitably come with building on a new platform. A worthwhile systems integrator should help you avoid those pitfalls by taking care of the heavy lifting while enabling a team's data engineers and data scientists to learn new skills along the way.