Refining Proximus’ AI-powered customer support chatbot
Proximus is paving the way when it comes to providing instantaneous support for both its customers and employees. They have been steadily developing a state-of-the-art chatbot powered by generative AI to assist with support queries.
Our services
UX Copywriting
Crafting clear, engaging and user-centric content to guide and inform users
UX Design
Creating efficient, relevant and pleasant experiences by designing individual user flows
What was the challenge?
Proximus is focused on further developing a customer support chatbot aimed at enhancing the user experience and operational efficiency. This technology enables them to address customer issues around the clock, allowing human agents to focus on more complex inquiries and less on repetitive tasks. However, in actuality, chatbots are a near-perpetual work in progress. As time went on and more people interacted with chatbot, more fringe cases popped up where the chatbot didn't know what to do or answered incorrectly. As a result, Proximus needed additional support in addressing these cases as well as creating brand-new flows.
What steps were taken?
- Need Detection: Identify new flow requirements.
- Analysis: Evaluate customer needs and technical implications.
- Design: Create user-centric & technically sound conversation flows.
- NLP Integration: Ensure accurate query routing.
- Technical Connectivity: Link flows to customer service for complex queries.
- Testing: Ensure smooth customer experience.
- Deployment: Launch in production.
What are the results
The chatbot already had a substantial foundation, as we were team of around 18 people in total. With our help, the team reached unprecedented numbers of resolved tickets through the chatbot, meaning more people were getting help without having to burden human agents.
What did we deliver?
Conversational Bot Implementation
Expanding existing conversational flows of the Proximus chatbot
Missing flows were first detected by the team and stakeholders, which we then referenced with customer needs and evaluated the technical implications as well. A conversation designer then got to work, supported by the NLP team to make sure new flows were triggered correctly. As a team, we thoroughly tested the new flows, added multi-language support, and finally put it into production. As of now, we're keeping a close eye on it–monitoring the reactions of users for future improvements.
Project skills used
Hard skills
These are the technical abilities and knowledge that were required to execute the task at hand. In this case, we needed:
Soft skills
Best defined as interpersonal attributes that make communication and collaboration work and run more smoothly:
Relevant expertise at every turn.
Looking for specialist skills or niche talent? We match you with our top talent, so you can build better in-house digital teams.
Let us know how we can help you
Looking for new talent, short or long term? Fill out this form and we'll get in touch with you shortly!
