How SAIL Interns Used ChatGPT to Build an Expense-Tracking App in Just Two Weeks


The past summer has been quite a fun ride for our Lab. Akib Sadmanee and I, had the privilege of hosting seven exceptional high school students in a unique computer science internship under Project Hōkūlani. And what these students accomplished in a short two-week period is remarkable.

Simplifying Complexity with AI When we started brainstorming ideas for the internship, one of the challenges was to pick a project that was ambitious but achievable. We decided to build a mobile app designed to help users track their expenses. Now, the exciting part of this isn’t just the app itself, but how we went about creating it.

In the past, breaking down receipt information like vendor names, phone numbers, and item lists would require heavy lifting from machine learning experts. But our highschool students used ChatGPT to perform these tasks with what’s known as “prompt engineering.” This significantly sped up the development process, allowing us to focus more on education and less on coding.

The students also utilized no-code platforms like Amazon Web Services to build various features such as user authentication and image uploading. And yes, ChatGPT was there to assist us with answering any technical questions that arose, making the process even smoother.

What strikes me the most is how this experience highlights the influence of AI technologies like ChatGPT on the future of software development. If high school students can build such a complex app in just 40 hours, most of which was spent on learning rather than coding, imagine the technological advancements we can expect in the coming years.

What’s next for next year interns? We’re not sure yet, but we’re excited to find out!