Storm Reply Helps Radarmeteo Transform its Geospatial Data Services Using AWS

Italy-based weather service provider Radarmeteo migrated to Amazon Web Services (AWS) and revamped its customer-facing apps and APIs, with the help of AWS Partner Storm Reply. With the scalability and reliability of AWS and the technical expertise of Storm Reply, Radarmeteo is now ready to continue its growth journey. Its new cloud infrastructure has helped the company to enter new markets following the project’s completion.

Radarmeteo is a meteorology company that provides professional weather forecasting services to organizations across Italy and beyond. Radarmeteo offers its customers—including insurance companies, energy firms, utilities, and airports—complete weather information, from historical weather data to forecast information.

To deliver weather data to clients, Radarmeteo offers a geographic information system (GIS) web application called MeteoCast. MeteoCast is an integrated weather support platform offering all Radarmeteo services and available weather data—both third-party data and Radarmeteo’s own data.

Radarmeteo’s clients can log into MeteoCast and pull out the reports they need, including charts, graphs, and tables. “Our forecasts and climate data are vital for our customers,” says Andrea Chini, chief operating officer (COO) of Radarmeteo. “Whether they’re getting complete forecasts through Radarmeteo and MeteoCast or accessing raw data via Hypermeteo—our new data platform—knowing how the weather will affect their operations is critical.”

But MeteoCast had limitations. The system was on-premises, used outdated technologies, and couldn’t grow as the company expanded. Radarmeteo was struggling to share complete datasets with customers.

The company turned to AWS Partner Storm Reply to redesign and refine the application to transform Radarmeteo’s customer service offering and change how it handles data behind the scenes.

Replacing its existing infrastructure was a priority. “MeteoCast was getting old,” says Chini. “It was hosted on-premises. The technology was very limited, and it was increasingly difficult for us to add new features and functionality.”

Radarmeteo also wanted to launch a whole new business, called Hypermeteo, which wouldn’t use MeteoCast. Instead, it would use API access to let customers integrate raw weather data within their businesses. However, its existing system greatly limited Radarmeteo and Hypermeteo’s ability to expand and offer new products and services.

Storm Reply has been an AWS Partner since 2013 and part of the AWS Well-Architected program since 2018. “Storm Reply helped us migrate to AWS and build a whole new application for Radarmeteo,” says Chini. “In addition, they engineered the APIs used by Hypermeteo to deliver raw data to customers.”

Storm Reply began developing the new serverless system in September 2020 to allow Radarmeteo to build and run applications without worrying about servers. The company’s cloud infrastructure was built using Amazon API Gateway, AWS Lambda, Amazon Aurora Serverless, and Amazon Elastic Container Services (Amazon ECS) instances. In addition, Radarmeteo uses Amazon Simple Storage Service (Amazon S3) for standard data services, plus Amazon Elastic Compute Cloud (Amazon EC2) for advanced data processing and radar models.

Front-end development started in March 2021 and go-live was in January 2022.

The company had ambitious expansion plans. “We are a small company of 25 people, but we are growing fast,” says Chini. “We needed the scalability of the cloud, and AWS offered the speed, power, and data handling capabilities we needed.”

Radarmeteo holds 30 years of historical data and offers a 15-day weather forecast. Naturally, this library of weather data is constantly expanding. The volume of new weather data handled by Radarmeteo grows at a rate of 15 GB a day. Some customers consume all of that every day, so the data throughput demands are high.

The company needed an infrastructure platform that could scale as the volume of data grows, while also delivering large amounts of data to customers daily. “We now store billions of weather values in AWS, around 100 TB of data,” says Chini. “It’s much easier to develop new features. What would have taken months to develop on the old system can now be completed in just a few days.”

The project has helped Radarmeteo deliver better services to its customers. The company has added new functions and now offers a whole new user experience that is faster and easier to use—a significant improvement on what the old system offered. “Radarmeteo can now offer its customers a greatly improved service through a completely overhauled application,” says Andrea Botta, manager at Storm Reply, who oversaw the project. “It can provide more weather data much faster, with improved regional specificity.”

Today, Radarmeteo’s customers can also create service tickets to check on the progress of their data requests. Customer data requests are easily created and amended—with the whole process fast and dynamic. “The previous on-premises system couldn’t manage the huge number of requests we received. Today, we are able to answer millions of data requests every year—all at speed,” says Chini. “Response times vary depending on the amount of data required, but it can be as little as just a few seconds.”

The new serverless system also gives Radarmeteo greatly enhanced scalability and elasticity of its infrastructure, made possible through the AWS serverless architecture. Radarmeteo no longer needs to spend money maintaining costly infrastructure, paying only for the server resources it actually uses. This keeps costs as low as possible. “The beauty of AWS is that Radarmeteo can tune the system to serve the company as it wishes without worrying about ongoing maintenance,” says Botta.

In addition to MeteoCast, Storm Reply has created a second application: a back-office system for configuring data related to each product or service and managing customer accounts.