Transformative software company achieves new heights in fraud detection
About Riskified
Founded in 2013, Riskified is a leading provider of fraud and chargeback prevention technology, operating as a software-as-a-service (SaaS) platform. Traded on the New York Stock Exchange (NYSE) and serving clients in over 180 countries, Riskified uses AI, machine learning, big data, and a deep understanding of consumer behavior to help large retailers and mobile delivery platforms like Macy’s, Wayfair, Prada, GrubHub, and DoorDash build secure and seamless payment solutions. Riskified not only helps businesses protect their revenue from fraud but also ensures that genuine customers don’t face unnecessary hurdles when they shop online.
Inefficient cost management when trying to balance fraud management and scalability
In a digital-first world, managing fraud has become an increasingly critical and complex part of doing business online. Riskified works to combat fraud by innovating software that removes risk and blocks fraud efficiently. However, Riskified needed to balance their customers’ need to conduct transactions seamlessly while having robust security to combat sophisticated fraud attempts without undue friction.
In order to do this, the company relied on real-time data to make decisions to support fraud detection, improve customer service, and enhance transaction accuracy. Riskified reviews transactions worth over $60 billion annually from more than 180 countries, creating a rich network of consumer history.
As the company expanded, several key challenges within their existing infrastructure emerged as they scaled their operations:
Data management
Efficiently handling large volumes of real-time data from diverse sources.
System scalability
Scale-up during peak traffic periods like holiday seasons, including Black Friday and Cyber Monday - without system strain.
Balance of speed and precision
Ensuring quick yet accurate fraud detection to prevent customer dissatisfaction and financial loss.
Cloud resource optimization
Efficient use of cloud services and managing technology licensing costs.
Enhanced scalability and performance with Aerospike
Riskified transitioned to Aerospike to address critical challenges, leveraging the technical solutions provided by Aerospike’s database to overcome key issues with their previous provider. The integration with Kafka significantly enhanced Riskified’s ability to handle real-time data, which is crucial for instant fraud detection. Additionally, Aerospike’s licensing model offered a cost-effective scaling solution, preventing proportional cost increases during traffic surges. The robustness of Aerospike ensured consistent reliability and improved the accuracy of Riskified’s data processing and fraud detection models.
Low-latency data storage
Aerospike’s Hybrid Memory Architecture (HMA) enabled vast data storage on SSDs with DRAM-like speeds for both read and write operations, essential for Riskfied’s instant fraud detection processes.
Seamless scalability
Aerospike’s horizontal architecture allowed limitless scale, efficiently managing traffic surges during peak periods. Efficient scaling during high-traffic periods
Real-time processing
Achieved real-time data processing with minimal latency
Improved fraud detection
Enhanced speed and accuracy in identifying fraudulent transactions
Significant gains in efficiency and cost savings
Riskified’s adoption of Aerospike led to significant benefits across both technical and business domains:
Increased loads from 2x to 8x
Aerospike was able to managed high-traffic periods without system strain, streamlined data management,and allocate resources to growth areas.
P99 at sub-10ms latency
The solution allowed for swift and reliable real-time data handling at submillisecond latency
99% customer retention
With enhanced fraud detection and transaction validation speed, Riskifeid delivered a more seamless customer experience.
Reduced costs
Riskified achieved significant cost savings through efficient scaling and maintenance.
Testimonials
We came to the conclusion that the previous technologies and architecture that we use will not scale and will not bring it to where we need it to be. And we started to look for solutions and happily found Aerospike as a good solution for those situations.
Shai Peretz
RiskifiedSVP of EngineeringAdditional resources
For a deeper understanding and more insights, explore these additional resources.
Get started with Aerospike
For high-performance, scalable data management and ultra-low latency, ideal for handling massive datasets and real-time applications.
