Project Reflection: E.G. Christmas Market Booking#
As part of my project, we had a meeting with E.G. about some of the challenges they experience in relation to their Christmas market. This meeting helped me understand the problem from a more practical perspective, because it was not only about building an AI solution. It was also about understanding what kind of support E.G. actually needs.
One of the main challenges is the process around booking stalls. When people want to book a stall, they often have many questions about prices, rules, deadlines, practical information and how the booking process works. Many of these questions are repeated by different customers.
This creates a lot of emails for E.G. It takes time to answer the same types of questions manually, and it can also make the booking process feel unclear for the people who are interested in having a stall at the Christmas market.
The problem#
The main problem is that information is not always easy for customers to find. If a customer has to send an email to ask about basic information, it creates extra work for E.G. At the same time, the customer has to wait for an answer.
Examples of repeated questions could be:
- How do I book a stall?
- What does it cost?
- What are the rules for stallholders?
- What deadlines should I know about?
- Who should I contact if I have a special question?
- What practical information do I need before the Christmas market?
These questions are important, but they do not always need a personal answer. If the information is presented clearly, many users can help themselves.
My solution#
Because of this, I have worked on a solution that combines a website with a chatbot. The website can present the most important information in a clear and structured way. The chatbot can then support users by answering common questions.
The goal is not to replace personal contact completely. Some questions will still need to be answered by a person. Instead, the goal is to reduce the number of repeated emails and make the first step easier for customers.
The website can work as the main information page, while the chatbot can work as a guide. A user can ask the chatbot about stall booking, practical details, rules and contact information. If the chatbot does not know the answer, it should not guess. It should guide the user to contact E.G. directly.
Why a chatbot makes sense#
A chatbot can be useful in this situation because many questions are predictable. If E.G. already knows the most common questions, those answers can be added to a knowledge base. The chatbot can then use that information to answer users in a consistent way.
This is also where RAG is relevant. A RAG-based chatbot can search in the knowledge base before answering. This means the chatbot can stay closer to the information it has been given, instead of making up answers.
For this type of project, it is important that the chatbot has clear limits. It should be able to answer general questions, but it should not approve bookings, promise available stall spaces or handle payments. Those tasks require more control and should stay with E.G.
What I learned#
Through this project, I have learned that AI does not always need to be used for very advanced solutions. Sometimes a simple chatbot connected to relevant information can already create value.
I also learned that the most important part of an AI project is not always the technology itself. It is also important to understand the real problem. In this case, the problem is not just “E.G. needs AI”. The problem is that customers need clearer information, and E.G. spends time answering repeated emails.
The combination of a website and a chatbot could help make the booking process more understandable, reduce repetitive communication and give customers a better experience when they want to book a stall for the Christmas market.
Reflection#
This project helped me connect AI with a real use case. It showed me that a useful solution does not have to solve everything at once. A small and focused MVP can still be valuable if it solves a concrete problem.
For E.G., the first version should focus on information and support. More advanced features, such as automatic booking, payment or admin systems, could possibly be added later. However, starting with a website and chatbot makes the project more realistic and easier to test.
Overall, I see this project as a practical example of how AI can support communication. The chatbot can answer the simple and repeated questions, while E.G. can spend more time on the questions that actually need personal attention.