Starting the project
Although I have familiarized myself with AI over the past few years through courses in my bachelor’s program, I had not used it beyond study and coding purposes. So, this assignment interested me since it allowed me to work with AI in another way that went further than simply using it for my own knowledge gain and simply generating textual outputs. The weekend before I started with the assignment I tried to book a football trip with some friends and experienced difficulty doing so because the websites we used were confusing. When we started brainstorming about a potential business case that would be suitable for the assignment my first thoughts were about how AI could be a solution for these unsatisfactory booking experiences.
My role in the project
My first job was to identify why current business processes caused so much friction and what the consequences of this could be for both the business and their potential customers. This way I managed to identify an area that would benefit from the intervention of AI which enabled us to design a targeted business case from which both the business and customers could benefit most. Applying multiple theories, such as Sweller’s Cognitive Load and Ainslie’s Present Bias showed that the businesses’ problem was the inefficient workflow that was currently forced on customers. Successfully identifying this allowed us to utilize AI in such a way that it solves the current inefficiencies effectively and I worked on explaining what benefits the use of AI could bring.
We did, however, encounter some uncertainties as the way AI would be implemented in our proposal would not create any significant new products and we were uncertain whether this topic would be ambitious enough. After discussing and thinking about potential other topics we did eventually conclude that this topic would be nice to research since it targets a really exact area that could directly improve and bring more value, both for the company and for the customer.
Once we established that this was going to be the topic for our assignment I worked on finalising the problem, what solution AI created and what the new business model would look like. I also made a mock dataset that contained data about football matches and trips across Europe which a teammate used to create our prototype. In hindsight we should have started setting up an A/B-experiment whilst waiting for the prototype, this way we could have started with the tests as soon as it was finished. However, I started setting up this experiment a bit later which caused a slight delay, not a big issue but something I will definitely remember for future projects. Setting up this experiment was interesting since it really forces you to think about the area that the business case targets and tries to improve. By doing this thoroughly the experiment and the data generated are complete without being too excessive and long.
Conducting the experiment and interpreting the results was rewarding to do since it confirmed the theories and the initial thoughts we had. Comparing the results from the old process and the workflow coming from our prototype showed that strain and annoyance declined when the user had all information present within one website. It is important to note that the prototype was a proof of concept and the number of participants was limited so real conclusions could not be made, but I still found this way of doing research and trying to prove assumptions an educational process that interests me.
What I enjoyed about the project
When forming our groups it became clear that all of us had different interests and strengths due to different academic backgrounds which was really helpful during this project which consisted of different aspects. This allowed us to divide the work in a way that most of our individual strengths were used whilst combining our ideas to improve the overall quality of the project. This was especially nice as this helped me understand how AI could be programmed and used in such a way that it solely does the activities that we need it to do.
Working on this assignment whilst simultaneously learning new concepts during the course’s lectures made me interested to deepen my understanding of AI and more specifically Agentic AI. When we tried to establish how AI could best serve our goal, which was improving the process of booking football trips, we addressed multiple different solutions. Discussing these topics and working on this assignment made me realize that there are a wide variety of applications for AI and that I understand quite a few of them but that I am rather unfamiliar with their exact application and how they can be implemented. Before this course and assignment I was well aware of the strategic power AI can have for a wide variety of companies but it is nice that this has become more concrete now.
