Owning Change in Communications
Our customer, a multibillion-dollar global tech titan, empowers their customers with end-user support. Their mission? Level up customer service across languages through automated self-service.
The Challenge: Taking Customer Support Global
Automating Call Center Support
Our customer sought to improve the overall support experience for end customers reaching out with inquiries about their various products and services. Due to their size and reach, our customer needed to make these improvements accessible to multiple international markets. By implementing virtual, self-service bots to serve customers across the globe, they could improve end-user experiences, reduce costs and free up human agents to support more complex requests.
- Create a standardized, end-to-end process to rapidly build bots in English and port them into international languages.
- Capture localized prompts lists and recognition grammars for international languages.
- Maintain high-quality delivery through comprehensive end-to-end testing, flow dev testing, hammer testing, compatibility testing and user acceptance testing (UAT).
Step one? Enlisting a team with Google Contact Center AI experience, international coverage and the ability to scale quickly. Enter TEKsystems.
Our Solution: Teamwork Across Time Zones
Developing Voice Bots and Chatbots
We were ready to answer the call. Working both stateside and from our solution centers in India, we activated a team of conversational architects, dialogue designers, bot developers, integration developers and QA engineers.
Together, we worked to design, develop and deploy multilingual voice and chatbots to support call center automation across the customer’s business units. Working on multiple bots in parallel, we designed voice user interface (VUI) specs for the voice bots and conversation flow specs for the chatbots for an enhanced conversational experience—meeting our customer’s gold standard in conversational design principles. The bots were designed to invoke fulfillment APIs to qualify, authenticate and route calls through routing bots, ID/auth bots and self-service bots.
Using Dialogflow CX, our team rapidly developed bots in multiple languages, including Portuguese, Spanish, Japanese, Korean, Hindi, French, Polish, Swedish and German. The bots also had region-specific localizations to detect the region-specific vocabulary, grammar and pronunciation, like English localizations for North America, Great Britain and Australia. These bot solutions were also trained to consider user behavior in each market to recognize intent more precisely. Measuring performance using tools like Chatbase and AAM 2.0, we consistently maintained an F1 score of 0.8 or above postlaunch for all bots.
Powerful Partnership: Speaking Your Language
Clear Communication and Collaboration
We worked with multiple businesses, third-party vendors and cross-functional teams—each with their own priorities, complexities and challenges. We focused on listening to truly understand the vision and needs of each team while adapting along the way.
To keep the project running smoothly, we held daily scrum meetings and weekly status meetings with stakeholders. This gave everyone the space to identify improvement opportunities and execute. Our close collaboration set the customer up for success and helped our teams learn and grow, too.
The developed voice bots and chatbots greet customers and capture intents through key phrases to ultimately qualify, authenticate and route calls to the appropriate product business unit. Through automation and self-service, customers can quickly find solutions or get connected to the right agents who can help them with more complex issues, avoiding typical support headaches.
The bots have supported a significant call center workload across multiple regions, markets and lines of business for the customer. By automating and empowering end users with self-service, our customer reached their call containment and call deflection goals. End-user support, realized.