Pi Securities delivers faster trading with cloud workloads in Thailand

As a securities broker serving primarily Thai investors, Pi Securities needed its cloud-based application platform to run closer to its users. With workloads hosted in the AWS Asia Pacific (Singapore) Region, investor requests were routed from Thailand to Singapore before connecting back to the company’s on-premises trading systems. This multi-hop network path increased latency and made it difficult to support time-sensitive trading workloads in the cloud, with features such as programmatic trading continuing to rely on co-location infrastructure.

“Some trading features require very fast execution,” says Nikun Khoongumjorn, head of technology, Pi Securities. “Because our cloud services were hosted outside of Thailand, it was difficult to support low-latency trading capabilities in the cloud.”

At the same time, Pi Securities sought to consolidate its environment on AWS after previously operating across multiple cloud providers, which introduced data residency considerations and regulatory ambiguity around keeping financial data onshore.  With the launch of the AWS Asia Pacific (Thailand) Region, the company saw a path to complete this consolidation and engaged DailiTech, an AWS Partner that had built and managed its cloud platform from the beginning.

Pi Securities worked with DailiTech to migrate its cloud-based application platform to the AWS Asia Pacific (Thailand) Region, bringing workloads closer to users and reducing latency for trading applications. Building on its knowledge of Pi Securities’ environment, DailiTech led the regional architecture design and migration strategy, mapping workloads across regions and adapting the existing infrastructure for deployment in Thailand.

To support the transition, DailiTech refreshed the AWS landing zone using updated architectures and infrastructure-as-code, and worked with Pi Securities through workshops and testing phases to plan workload sequencing, data synchronization, and cutover. This approach reduced migration risk while maintaining regulatory requirements and operational continuity.

Pi Securities migrated non-production workloads first to surface issues early, before running production workloads in parallel across both regions ahead of full cutover. “Running production workloads in parallel across both regions gave us the confidence to complete the cutover without disrupting our investors,” says Nikun Khoongumjorn.

Pi Securities runs its application platform on Amazon Elastic Kubernetes Service (Amazon EKS), Amazon Elastic Compute Cloud (Amazon EC2), and Amazon Relational Database Service (Amazon RDS), now deployed in the AWS Asia Pacific (Thailand) Region, enabling the company to process investor requests locally while maintaining connectivity with its on-premises trading systems. It uses Amazon Simple Queue Service (Amazon SQS) and Amazon Simple Notification Service (Amazon SNS) to manage communication between microservices, allowing requests and trading events to flow reliably without bottlenecks. The platform also uses services such as Amazon CloudWatch, AWS CloudTrail, and AWS Secrets Manager for monitoring, audit logging, and secure configuration management.

Pi Securities reduced application latency from 100 ms to 10 ms—a 10x improvement that enables the company to support latency-sensitive trading features in the cloud. Faster response times now allow Pi Securities to develop capabilities such as programmatic trading without relying on co-location infrastructure.

The consolidation delivered a 15 percent reduction in infrastructure costs, with all cloud workloads now running on AWS in Thailand to meet data residency and regulatory requirements. Running workloads locally also gives the engineering team greater flexibility to build new features without infrastructure location as a constraint.

"With infrastructure now running in Thailand, we can build programmatic trading capabilities in the cloud that simply weren't possible before. That's a significant competitive opportunity for us," says Nikun Khoongumjorn. Building on this, Pi Securities plans to explore Amazon Bedrock for customer-facing AI features and Amazon Q to help its sales teams surface actionable insights from customer behavior data.