Gamota improves support operations with ECVBot, reducing manual workload by 50%

As Gamota grew its presence across Southeast Asia, its support team faced rising complexity in assisting customers, who are primarily game players, across multiple countries with distinct languages, cultures, and time zones. These differences made it challenging to deliver consistent, high-quality responses from a central team.

Player engagement also increased across Gamota’s portfolio of titles, leading to higher inquiry volumes and more questions that required detailed context, such as purchase history, in-game progress, event information, and troubleshooting. Managing this workload slowed response times and placed additional pressure on the team.

To provide fast, accurate, and multilingual support across the region, Gamota needed a more scalable and efficient solution. Kevin Duong, chief technology officer at Gamota, explains, “With more than a hundred active game titles, every title brings different issues, and managing that level of variety and detail across the region meant rethinking how we approached customer support.”

With core systems such as its Customer Relationship Management (CRM) and game services already on AWS, Gamota engaged AWS Partner eCloudvalley to deploy ECVBot, a generative AI chatbot powered by Amazon Bedrock. eCloudvalley worked with Gamota to customize the solution so it could support the way players interact with the company’s game mechanics and internal workflows.

The two teams coordinated daily through shared communication and task management channels. Both teams worked through these channels to refine features, address issues, and make iterative improvements quickly. “As an AWS Partner, the eCloudvalley team always took the time to examine each issue carefully and understand what was really happening before proposing changes,” Duong says.

Gamota and eCloudvalley integrated ECVBot with Gamota’s internal systems, such as player purchase history, ticket records, and in-game data. These integrations helped the chatbot generate contextual summaries for support teams and interpret player questions with greater accuracy. To improve both speed and reasoning quality, the teams used Amazon Bedrock to work with multiple foundation models, selecting faster models for simple queries such as retrieving account information, and reserving larger ones for cases that required deeper reasoning or data lookups, such as investigating purchase history or payment-related issues. “With Amazon Bedrock powering ECVBot, we could choose the right model for each type of inquiry. That made the chatbot faster, more accurate, and easier to configure for our games,” Duong says.

Through this collaboration, ECVBot was fine-tuned using Gamota’s game data and support scenarios, making it capable of handling a wider range of inquiries and adapting to the nuances of different mobile games and languages. The iterative approach also helped Gamota keep pace with the speed of development cycles, enabling the chatbot to be continually updated as new weekly or fortnightly game events, features, and content updates are introduced.

Since deploying ECVBot, Gamota has been able to manage rising inquiry volumes of 15–20% more effectively while maintaining consistent support quality across its portfolio of more than 100 mobile games. By handling routine and commonly asked questions, the chatbot has reduced the number of cases that require manual review. As more simple inquiries are addressed automatically, Gamota’s support teams can focus on resolving complex player issues, reducing overall human agent workload by 50 percent.

The platform has also strengthened multilingual support by automatically detecting and responding in the user’s original language. Players can ask questions in Vietnamese, Thai, or Bahasa Indonesia, and the chatbot replies in the same language—eliminating the need for native-language agents and ensuring consistent service quality across markets.

Beyond multilingual responses, ECVBot also generates concise player information summaries in under 30 seconds, helping support teams understand each case more quickly before taking action. With faster case understanding and automation in place, about 90 percent of inquiries are resolved in minutes, giving players a faster and more consistent support experience across markets.

In addition, Gamota is able to monitor recurring issues reported by players. ECVBot helps surface patterns in the types of questions being asked—for example, several players encountering the same gameplay obstacle or raising similar payment concerns—which helps the team notice these trends earlier and begin looking into them promptly. “ECVBot helps our team understand issues earlier and react more quickly when many players are affected,” Duong says.

These improvements support Gamota’s plans to expand its use of generative AI into areas such as game recommendations, creative content generation for marketing, and other future use cases that will help enhance player engagement and operational efficiency.