Zego Drives Down Insurance Premiums and Keeps Roads Safer Using AWS

Zego provides specialist insurance for short-term drivers such as private hire and delivery drivers and couriers. When it acquired driving telematics company Drivit, it needed to rapidly and securely migrate Drivit’s systems from another cloud provider onto AWS. Using AWS, Zego met a tight migration deadline and moved 60 TB of data without impacting customers. It’s now able to scale to meet variable demand for its services and use machine learning (ML) to support competitive pricing and improve driver safety.

Zego provides specialist, short-term insurance for private hire anddelivery drivers, and couriers. It offers flexible, hourly, or daily plans, providing cost-effective options for shift workers.

The company has been all-in on Amazon Web Services (AWS) since it sold its first policy in 2016. When it acquired driving telematics company Drivit, it needed to rapidly and securely migrate Drivit’s systems from another cloud provider onto AWS. The move avoided the expense of running two parallel systems.  

Using AWS, Zego met a tight 9-month migration deadline and moved more than 60 TB of data without impacting customers. It’s now able to scale to meet variable demand for its services and use machine learning (ML) to support competitive pricing. It also analyses data on driver performance to  help customers lower their risk profiles and educate their drivers to improve road safety.

Migrating 60 TB of Data to AWS on Time and under Budget

Drivit monitors drivers’ performance constantly to assess safety, collecting 50 data points every minute, including driving speed, acceleration rates, and braking. This created huge quantities of information for Zego to migrate to AWS. 

The team used AWS Database Migration Service and CLOUDBASIX for RDS SQL Server to migrate five Microsoft SQL Server and PostgreSQL databases to AWS. It also moved tens of millions of files (60 TB of data) into Amazon Simple Storage Service (Amazon S3), an object storage service, so that the data could be classified. The migration went smoothly, finished within the deadline, and came in under budget. 

Zego worked with AWS Partner DevOpsGroup to ensure an easy transition. “DevOpsGroup has great engineers and developers, and they took the time to understand us and what we needed,” says Bart Swedrowski, director of systems engineering at Zego. “They led the project and drove it forward. They were also flexible when plans and approaches changed.” 

Scaling to Meet Rising Customer Demand and Cutting Costs by 50%

Zego insures drivers by the hour and day, but also offers monthly and annual policies. This means its systems experience variable demand, depending on how many drivers are working at any one time. These spikes were exaggerated during the COVID-19 pandemic, when uncertainty over lockdowns made short-term insurance attractive to more drivers and demand for delivery services increased. 

Using AWS, Zego can easily adjust resources to cope with fluctuating demand for its services.It stores data in Amazon S3 buckets and runs on Amazon Elastic Kubernetes Service (EKS), a fully managed service to run and scale Kubernetes applications in the cloud or on premises. “Running on EKS gives us the scalablity and reliability we need,” says Paolo De Luca, senior engineering manager at Zego. “We’ve done the stress tests and are confident we can handle any loads we have to deal with, even during our busiest times.” 

Being able to adjust compute resources up or down based on demand, coupled with better AWS performance, means that Zego has reduced compute costs by 50 percent on AWS, compared to its previous setup. Applications running on Amazon S3 also perform up to 80 percent faster than with Zego’s previous provider.

Driving Down Premiums Using Real-Time Analysis and Machine Learning on AWS

Zego can now analyze driver data in near-real time, using machine learning. It uses Amazon SageMaker to build, train, and deploy machine learning models, and Kubeflow on AWS to build sophisticated machine learning systems.  

Monitoring real-time information about driver performance means that Zego can better control its risk profile. It can dynamically price insurance based on actual driving performance and safety, and provide competitive rates, with lower premiums for safe drivers and higher premiums for drivers with consistently poor performance. 

For customers managing large fleets of vehicles, Zego can provide almost instant feedback on driver performance to improve safety. When systems detect anomalous driving behavior, they alert customers by text message so that fleet managers can ask drivers to change their behavior. This keeps roads safer and helps users avoid paying higher premiums. 

Zego can also offer competitive premium rates to fleet managers. Using AWS, Zego assesses risks on a driver-by-driver basis and provides a risk profile for each. This differs from traditional insurers, which provide a single risk profile for the entire fleet. This means that Zego customers can reduce premiums by training the riskiest drivers, who would otherwise push up costs. 

Fleet managers can also use data from Zego to help reduce the environmental impacts of vehicles—for instance, by optimizing routes so drivers use less fuel.

Ready to Innovate for the Future Using AWS

Zego is always looking for ways to make better use of AWS to improve data analysis and customer services, and to optimize costs. To reduces storage costs further, the company plans to move its data to Amazon S3 Glacier, which provides long-term, secure, and durable storage classes for archiving.

Since unifying Drivit’s systems with its own on AWS, Zego’s infrastructure is more reliable, agile, and simple to operate. It also supports future innovation. “Using AWS, we have everything we need to push forward and stay ahead in the fast-moving insurance business,” says Swedrowski.