Qiscus Empowers Websites with Chat SDK Built on AWS
Real-Time Customer Service
Real-time communication (RTC) is increasingly becoming a staple of successful companies’ customer service offerings. Qiscus, a software vendor based in Indonesia, presents two options for live chat platforms: a software development kit (SDK) to enable conversations on a company’s website and mobile app; and a multichannel chat service that customers can use to chat with companies via WhatsApp, Facebook, and other social media messaging applications located on a single dashboard.
When setting up its SDK in 2016, Qiscus evaluated cloud providers, including Amazon Web Services (AWS), based on reliability, cost, and their integration with other infrastructure layers such as databases. “AWS was the most stable and predictable in terms of latency and performance. None of the other providers had such consistent service,” says Evan Purnama, cofounder and chief technology officer of Qiscus. Since launching on AWS, the company has maintained an end-to-end latency for customer-facing APIs of 100 milliseconds or less.
Qiscus maintains a 99.9% SLA while growing substantially.
-
About Qiscus
-
Benefits
-
AWS Services Used
-
About Qiscus
-
Qiscus is a startup based in Indonesia that helps companies connect with customers using real-time communication technology. Its two main products are a chat software development kit that enables conversations on a company’s website and mobile app, and a multichannel chat platform integrated with Facebook, WhatsApp, and other social media messaging apps.
-
Benefits
-
- Maintains consistent end-to-end latency of 100 milliseconds or less
- Scales to accommodate rapidly expanding user base
- Ensures 99.9% availability for customers
- Provides automated, integrated monitoring tools
- Controls costs with visibility into per-service spend
-
AWS Services Used
Automated Monitoring Tools
Monitoring is vital to ensure Qiscus’s platforms maintain the latency required for RTC. Engineers developed their own monitoring system a year after launching, but before then relied on Amazon CloudWatch to send performance alerts. “Using Amazon CloudWatch gives us a lot of visibility into our servers. Whenever there’s a spike or drop in latency, we always know the exact cause,” Purnama says. The company relies on Amazon ElastiCache to scale and maintain caching functions for almost all applications and Amazon Elasticsearch Service to index and quickly retrieve aggregated data from its servers.
To integrate Qiscus’s multichannel chat service with its clients’ customer relationship management (CRM) applications, the business chose AWS Lambda, which also serves search results that are generated by Amazon Elasticsearch Service. “AWS Lambda automatically scales and integrates with Amazon CloudWatch to help us find and repair any problems in our code,” Purnama adds.
Achieves SLA of 99.9%
In 2017—Qiscus’s first year selling its chat SDK—Qiscus’s customer base reached 1 million users. This figure then rose to 5 million users in 2018 and exceeded 17 million users in 2019. Now, the company’s goal is to acquire 25 million users by the end of 2020. Despite this substantial growth, Qiscus has maintained a service level agreement of 99.9 percent—a feat Purnama partially attributes to the ability to utilize AWS servers in multiple Availability Zones.
Cost control remains a priority for the startup, because its customers can range from multinational enterprises to home-based operations that have a limited budget for chat services. When it comes to monitoring costs, Qiscus was very impressed with AWS Cost Explorer, which breaks down how much each service costs and tags individual services to analyze its monthly spend. “This enables us to make more informed decisions when structuring our pricing models,” Purnama explains. Looking ahead, Qiscus plans to explore machine learning on AWS as part of its efforts to continuously innovate its core products.
The diagram below illustrates Qiscus’s AWS infrastructure for its chat SDK and other RTC-based solutions:
Learn More
To learn more, visit Cloud Compute with AWS.