Enables low-latency, natural voice interactions for call centers with Gemini Live API
Helps automate 76% of incoming calls, realizing higher customer service quality
Supports instant and accurate customer service chat responses through Gemini 2.5 Flash
Reduces AI infrastructure maintenance workloads with Google Kubernetes Engine
Telexpress provides natural AI call center services with the Gemini API, enabling higher productivity and service quality.
Telexpress provides natural AI call center services with the Gemini API, enabling higher productivity and service quality.
Building human-like automated call center services with generative AI
Customer service automation has become increasingly common. Merging AI and human expertise opens possibilities to make customer service operations more efficient and flexible, leading to higher customer satisfaction.
With more than 25 years’ expertise in customer relationship management (CRM), Telexpress has been at the forefront of customer service modernization. Currently serving more than 100 leading brands across industries, it provides an AI platform that aggregates SaaS solutions covering all aspects of customer experiences, including voicebots, chatbots, agentic workspace and conversational CRM. Katrina Chen, product strategy and marketing manager at Telexpress, notes that the arrival of generative AI has shifted Telexpress’ product development focus from CRM modernization to customer experience optimization.
We wanted to build AI voicebots that react and sound like human customer service agents. Google Cloud was our go-to option, because Gemini Live API supports excellent voice interaction quality, and Google Cloud holds key security certificates to provide high information security assurance to our enterprise customers.
Katrina Chen
Product Strategy and Marketing Manager, Telexpress
“In the past, most of the AI customer service solutions were created to answer repetitive questions and reduce costs. But with generative AI, these solutions can go beyond answering questions to serve as crucial first contact points between companies and their customers, creating business value,” she explains.
Before generative AI matured, Telexpress used speech-to-text, text-to-speech technology and flow-based conversation engines to build interactive voice response (IVR) systems. With their menu-driven design and long latency, these systems provided slow and rigid responses. Callers needed to follow fixed procedures to retrieve the answers they wanted, resulting in poorer user experiences than human customer service agents could provide. Due to the technical restrictions, their self-service capabilities were also limited. To improve its voicebot solutions, Telexpress decided to employ generative AI and chose Google Cloud for the outstanding voice interaction performance of Gemini Live API in Gemini API and the advantage of Google Cloud in security, stability, and global presence.
Enhancing customer experiences and customer service efficiency with Gemini Live API
In April 2026, Telexpress integrated Gemini Live API into its one-stop customer service SaaS platform to support the AI voicebot solution. This integration makes it possible to get the latest information through retrieve-augmented generation (RAG) and understand users’ intent by intent-based routing customized for each enterprise customer. Telexpress’ enterprise clients can also implement their own customer service workflow rules, such as which tasks should be handed over to human agents and when to transfer. This way, the Telexpress AI voicebot is able to quickly grasp what callers want and provides relevant, accurate replies, all following each company’s requirements.
The advanced natural language understanding and generation capabilities enabled by Gemini Live API have greatly upgraded our voicebot solution. Now, our enterprise customers can provide natural, efficient, automated call center services, freeing their staff from repetitive work and enhancing customer experiences.
Katrina Chen
Product Strategy and Marketing Manager, Telexpress
For end users, the most remarkable difference between Gemini Live API and the legacy IVR systems is the conversational fluidity. As Gemini Live API enables around 60% lower latency, callers now enjoy human-like response speed. With features that create natural conversations, such as voice activity detection (VAD), which allows a user to interrupt the AI at any time, Gemini Live API enables smooth, empathetic real-time interactions. Callers can now express their needs in their own ways, instead of following fixed instructions, leading to higher conversation quality.
The results are also reflected through numbers. Save & Safe, a Taichung-based hypermarket chain in Taiwan, employs Telexpress’ AI voicebot to handle all retail stores’ incoming calls and sees 76% of calls resolved before being handed over to human staff.
Before the adoption, Save & Safe’s employees at each store were constantly interrupted by calls while trying to do their work. Now, with 40 calls reduced per store per day, they’re able to focus on more complex cases and their core tasks, providing better customer services both on the spot and over the phone.
Creating business value through proactive AI-driven customer interactions
Besides AI voicebots, Telexpress also leverages Gemini 2.5 Flash models to enable its AI chatbot solution. Together with other generative AI solutions, Gemini models provide instant and accurate text replies, resulting in a high self-service success rate. For example, an international cosmetics group adopted Telexpress’ chatbot solution to support its brands’ customer services and raised the direct response rate of AI from 70% to 87.7%, with more than 400,000 AI-enabled interactions per year.
“For us, the value of AI is more than answering questions. By understanding customers’ needs, connecting to companies’ internal systems to complete business tasks and collaborating with human agents, it can make every interaction with customers become enterprises’ key asset for continuous service optimization,” explains Chen.
To reduce infrastructure management workloads, Telexpress uses Google Kubernetes Engine (GKE) to deploy its AI solutions. The autopilot mode of GKE automates Kubernetes cluster management, allowing Telexpress’ DevOps team to spend more time on product innovations.
“What we like about Google Cloud is that it not only offers advanced AI capabilities, but also robust infrastructure that helps boost productivity,” says Chen. “For a company that aims at quickly developing AI SaaS solutions like us, Google Cloud serves as more than a cloud platform. It’s the foundation for the continued evolution of our products.”
The customer service sector is going through a massive transition with the growth of AI adoption. Using Google Cloud, we successfully upgraded our voicebot and chatbot solutions to enable higher productivity and service quality. As the technology of Google Cloud advances, we’re looking forward to further transforming our AI-driven customer services for greater business benefits.
Katrina Chen
Product Strategy and Marketing Manager, Telexpress
Telexpress is constantly updating its tech infrastructure by analyzing past customer interaction data and using it to train AI models. To provide services that meet real needs, the company will continue transforming its AI customer service solutions, adding more features proactively, such as product recommendations, sales support, and business opportunity identification.
“We believe that one core value of AI customer service solutions is to drive continued business growth, and that’s why we’re developing solutions that create business value,” notes Chen.
Founded in 2001, Telexpress provides enterprise-grade AI interactive services and customer service solutions. Focusing on the Asia-Pacific market, it currently serves more than 100 leading brands across industries.
Industry: Technology
Location: Taiwan
Products: Gemini 2.5 Flash, Gemini API, Gemini Live API, Google Kubernetes Engine
