Safehub Transforms Its Database on Amazon Aurora and Sees 80% Drop in Code

Safehub is transforming catastrophic risk management to improve business continuity through near-real time data assessment and communication. Its platform applies analytics to IoT sensor data, structural dynamics, and seismic activity to provide companies with swift assessments of property damage immediately following a potentially catastrophic event. Safehub worked with ClearScale, a cloud services company, to leverage Amazon Aurora, increasing speed, reliability, and security in its operations and data management.

When Safehub initially began collecting data to track seismic activity, it captured both relational and non-relational data, using PostgreSQL tables to organize some of the information. As the company honed its IoT model for predicting building damage in the event of an earthquake, it reached a point where coding complex queries for non-relational data was slowing data processing as well as communication with customers.

The company needed a fast, reliable, stable, and scalable database to ensure property damage insights reached customers in near-real time during an event. Obtaining more precise query results faster would also unlock opportunities for Safehub to scale and serve new customers.

In evaluating its options, Safehub chose to consolidate and migrate its databases from Amazon DynamoDB and Amazon Relational Database Service (Amazon RDS) PostgreSQL to a single Aurora instance, gaining scalability, speed, and enhanced data security across development, staging, and production environments. Because Aurora is compatible with PostgreSQL, Safehub could easily transition its existing PostgreSQL tables without rewriting the data model. The team could utilize SQL queries to access its datasets, in addition to taking advantage of auto-scaling and AWS Availability Zone redundancy. To increase security, Safehub chose to implement an Amazon Virtual Private Cloud (Amazon VPC) combined with a Virtual Private Network login to control access to customer data.

“Being all-in on AWS is a gamechanger,” said Andy Thompson, CEO of Safehub. “With Aurora in a VPC, we’re able to give critical earthquake damage information to corporations within minutes, so they know the status of their facilities as quickly as possible.”

The Safehub team knew where it wanted to go, but finding the most efficient path to get there without disrupting its ongoing work clarified the need for a qualified partner. Cloud services company ClearScale delivered the insightful guidance and precise execution Safehub needed.

ClearScale set up a new Aurora environment in an Amazon VPC, used an Amazon Elastic Compute Cloud (Amazon EC2) platform to consolidate data, helped clean non-relational datasets to fit within a relational structure, and refactored the solution for IoT optimization. The Amazon VPC provides an extra layer of protection for all of Safehub’s data, and the team is able to carefully control access points and work in an environment free of sensitive customer data.

“Amazon Aurora made it easy for us to find the perfect solution for Safehub’s needs,” said Sergey Krasnov, Solutions Architect at ClearScale. “With EC2 and SQL transformation commands, we were able to swiftly transition their databases to a higher performing, safer system in Aurora.”

ClearScale also architected a responsive backend that laid the groundwork for CI/CD pipelines to Aurora so Safehub could easily migrate its data. Before, Safehub faced many manual tasks to push new features to production. With ClearScale’s solution, Safehub can deploy with more agility.

“ClearScale helped us completely redo our whole AWS environment, incorporate our IoT data, migrate our database to Aurora, and set up our VPC,” said Alex Mead, Manager of Technology at Safehub. “They gave us a lot of confidence in our decisions and really brought value in getting the scripts right the first time.”

Compared to performing the migration itself, Safehub was able to migrate to Aurora two times faster with the help of ClearScale. And consolidating Safehub’s database deployments on Aurora helped to reduce operational costs by approximately 50 percent.

The move to Aurora significantly improved Safehub’s ability to query all data, especially geospatial data available from the United States Geological Survey. Moving to a relational database structure allowed a vast reduction in code complexity. Aurora enabled the Safehub team to quickly comb data with simple queries, streamlining its processes, easing data joins, and returning answers much faster with less effort. Safehub realized an 80 percent reduction in code, which translates to an easier, faster time to insights for customers.

“We need to quickly understand if there is a problem, what damage may have occurred, and what the response should be,” Mead said. “With Aurora, we can provide that information in near-real time, in less than two minutes to customers, so they can prioritize impacted areas and implement disaster recovery strategies.”

“When we have our buildings equipped with Safehub, we can immediately make a decision as to what the occupants should do,” explained Ricardo Gianvito, Head of Systems Analytics and Architecture at Amazon Corporate Security. “This kind of system is valuable to protect our Amazonians and know whether they should evacuate a building or shelter in place.”

Together, ClearScale and Safehub migrated hundreds of thousands of rows of data, including 25 building portfolios with approximately 5,000 buildings and 30,000 earthquake events. The data is now ready for queries to help Safehub track building safety and disaster recovery across a growing customer base.

“We are very happy with Aurora—and we’re ready to scale,” Thompson said. “AWS is going to help with that, and ClearScale was instrumental in making this migration a success.”