The Snaphot
The existing chatbot was limited providing generic responses, and the call center was overloaded
Usage of Generative AI to build a new chatbot, the first transactional Gen AI chatbot in Europe
Better customer satisfaction, with a more dynamic, transactional and self-service platformrn
The Challenge
CTT – Correios de Portugal, S.A. is a Portuguese customer-oriented company, with offers for companies and individuals. Its solutions vary from mail post services, order and package deliveries (including express services), saving offers and a broad portfolio of business products and services, namely e-commerce, logistics, advertising, digitalisation and process outsourcing solutions. CTT has a unique history of dedication, commitment and professionalism, which has brought the Portuguese and the world together for several centuries. The company was founded in 1520 by D. Manuel I of Portugal, making it the oldest company still operating in Portugal.
CTT faced a crucial challenge in an era demanding rapid and accurate customer service: modernising their support channels to improve efficiency, and quality of customer interactions. Their existing menu-based chatbots were limited and maintenance-intensive, failing to meet the evolving needs of consumers. This inefficiency was compounded by a surge in call volumes at their customer service centre – AVC, with many calls concerning queries that could be easily resolved digitally.
The existing chatbot system was limited, providing only generic responses that didn’t effectively resolve customer queries, leading to customer dissatisfaction and an overburdened call centre.
Given the context, CTT determined the following objectives: to reduce the call centre load by diverting routine inquiries to more advanced digital platforms, and to enhance the speed and accuracy of customer responses. To achieve this, the aim was to transform the service via chatbot through a new platform, more dynamic, transactional and self-service platform, thereby improving customer experience and operational efficiency.
“One particular kind of contact that sticks out at CTT is the object’s location. Finding out the location of the item, when it will be delivered, etc., accounts for 50% of interactions with CTT. We set a requirement for Devoteam that included having a transactional use case. ”- Diogo Freitas de Oliveira, Transformation Director at CTT.
The Solution
Devoteam has revolutionised CTT’s approach to customer service by developing “Helena”, an advanced chatbot powered by Generative AI, using Microsoft Azure OpenAI Services technology.
Designed to simulate human-like interactions, “Helena” leverages advanced AI to engage in natural language conversations, providing a seamless and efficient customer experience. This strategic implementation ensured consistency in customer service and introduced the capability for customers to execute self-service tasks effortlessly.
Helena was crafted to synthesise data from various sources via APIs, offering tailored, conversational responses. This integration allows for a robust response system based on CTT’s website and internal data, ensuring accuracy and a personalised touch in every interaction. Special care was taken to align “Helena’s communication style with CTT’s standards, embodying the friendly and informative nature of CTT’s human customer service, thus maintaining the company’s commitment to excellence in customer engagement.
“Helena’ has a CTT tone of voice and that was the great work we had throughout the project and where the greatest effort of the CTT team and Devoteam was focused. We changed every component of the prompt engineering, ensuring not only the correctness of the response content, but that the tone of voice was aligned with the CTT tone. Probably 70% of the project effort was making sure this tone was right.” – Carlos Machado Silva, Head of Corporate Transformation & Intelligent Process Automation at CTT
The Benefits
The deployment of “Helena”, developed by Devoteam, has significantly enhanced CTT’s customer service, marking a pivotal advancement in their service capabilities.
“We do not distribute packages, objects, or orders, we distribute emotions and as such, the customer is very eager for information and interacts with us a lot to know almost instantly where their order is. We felt the need to make this update.” – Cláudia Serra, Director of Customer Support and Voice at CTT.
This transformation is quantitatively evident in the leap of CTT’s Net Promoter Score (NPS) increase of 40 points, a direct reflection of the chatbot’s positive impact on customer satisfaction.
Key Achievements:
- Enhanced customer service: “Helena’s AI-driven conversations have personalised and streamlined customer interactions, making service more efficient.
- Operational efficiency: The chatbot has reduced reliance on call centres by efficiently handling a broad spectrum of queries, thereby lowering operational costs.
- Improved customer satisfaction: The substantial increase in NPS highlights the successful enhancement of customer experiences and validates the strategic approach of this project.
Transformative Results:
- Reduced call center volume: “Helena” has significantly decreased call centre traffic by providing fast and accurate responses to common inquiries.
- Agile information delivery: Integration with multiple data sources ensures Helena offers up-to-date and precise information, improving service quality.