Challenge
Wakefit faced the challenge of scaling its customer service operations to handle increasing demand while maintaining high-quality, contextually relevant interactions. With a wide range of products and services, the company needed a chatbot to offer personalized, real-time responses to diverse customer queries. Wakefit also required a system that could scale without compromising performance while ensuring that all data was processed efficiently to deliver accurate and relevant answers. The complexity of integrating multiple data sources into a single AI-powered interface was a key concern, along with the need to meet strict data residency requirements in India.
Solution
To address these challenges, Wakefit partnered with AWS and DataStax to build a GenAI chatbot solution. They chose Astra DB in DataStax’s AI Platform to handle massive amounts of structured and unstructured data with low-latency performance. Astra DB's native integration with vector search enabled Wakefit to create vector embeddings of customer data, allowing the chatbot to deliver highly relevant responses by retrieving similar customer queries from past interactions.
Amazon Bedrock provided Wakefit with the GenAI infrastructure, allowing them to integrate large language models (LLMs) into their chatbot system for natural language processing (NLP). The LLMs powered by Bedrock processed and understood the nuances of customer queries, while Astra DB ensured that the necessary data was instantly accessible for generating relevant responses. Astra DB’s scalability and fault tolerance were essential to handling high-volume customer queries, especially during peak hours, while ensuring zero downtime. Wakefit also leveraged Astra DB’s built-in multi-region capabilities, ensuring data residency compliance within India, a critical requirement for their business.
"We chose Astra DB and AWS because of their seamless integration and ability to scale with our needs. Astra DB’s low-latency performance and vector search made it the perfect choice for powering our GenAI chatbot. AWS Bedrock allowed us to implement and deploy advanced AI models quickly," said Puneet Tripathi, Head of Data Science at Wakefit.
Impact
The GenAI chatbot has significantly impacted Wakefit’s customer service capabilities. With the combined power of Astra DB and Amazon Bedrock, Wakefit’s chatbot now processes customer queries with near-real-time responses, drastically improving customer satisfaction. Using vector embeddings in Astra DB has increased the chatbot's accuracy in delivering contextually relevant answers. The solution's scalability ensures seamless operations even during traffic spikes.
Thanks to the rapid deployment capabilities of Astra DB and AWS, the transition from experimentation to a fully operational chatbot was completed in just 10 days. Wakefit's operational efficiency improved as the chatbot offloaded routine queries from the customer service team, allowing them to focus on more complex customer interactions. Additionally, the solution has provided Wakefit with a flexible architecture that can quickly scale as they introduce new product offerings and expand their services.
"The speed at which we went live was remarkable—just 10 days! The GenAI chatbot has dramatically improved our customer service, allowing us to handle more queries more accurately while our team can focus on complex issues. It has been a game-changer for our operations," said Mr. Puneet Tripathi.
