Who is using VeloDB?
TrustDecision is a decision intelligence solution that leverages AI and data science to identify frauds and credit risks for companies. Headquartered in Singapore, it has subsidiaries in Indonesia, Malaysia, and the UAE. So far, it has a user base of over 300 enterprises worldwide from industries including digital finance, e-commerce, electronic payment, banking, aviation and travel, media, and crypto currency.
How do they use VeloDB?
TrustDecision employs VeloDB in its data reporting and analytics platform. It integrates data from devices, customer business systems, and past risk management decisions, which together constitute a data size of almost a billion rows. Such data is collected by Apache Kafka and transferred to VeloDB, which works as the analytic engine, for keyword searches, detail queries, real-time reporting, and aggregated analysis.
How are they benefiting from VeloDB?
Before VeloDB, TrustDecision used Elasticsearch as the analytic and storage engine, and encountered some unbearable issues:
- No support for cross-table joins, and limited support for complicated aggregated analysis, not friendly for data analysis at scale
- High operating cost, no support for storage-compute separation, no elastic scaling
- High maintenance cost, and inconsistent user experience. With data centers across Asia, North America, and Europe, TrustDecision needs to guarantee consistent performance in all environments.
After a thorough research, they replaced Elasticsearch with VeloDB, which is now available on AWS, Alibaba Cloud, and Huawei Cloud. VeloDB fills in what was missing to make an efficient and scalable data analytic solution.
- Real-time performance: With VeloDB, TrustDecision achieves a writing throughput of 1000s of TPS and provides real-time query response for a billion-row data size.
- Usability: VeloDB is a SQL-based query engine, so it is friendly to most data engineers and analysts. It provides inverted index to make keyword searches a bliss.
- Elastic scalability: VeloDB is built on a cloud-native architecture. With storage and computing components separated, it allows clusters to scale out and scale in quickly. That can bring a 300%~500% increase in resource efficiency.
- Multi-cloud support: VeloDB runs on multiple cloud platforms. By working closely with cloud service providers, it ensures smooth user experience in various cloud environments. The fully-managed services also save a lot of trouble in operation and maintenance.
Based on the successful experience with VeloDB, TrustDecision plans to explore wide usage of VeloDB in their use case, including log management and analysis, metrics tracking, risk control and monitoring, and data lakehouse. They also plan to bring the solution to more cloud platforms and regions. We look forward to a deeper collaboration with TrustDecision and will keep enhancing our analytic solution to empower their business.