I might be a bit long-winded, but I really want to hear a lot of your thoughts, so…
- User
I might be a bit long-winded, but I really want to hear a lot of your thoughts, so please keep talking. I'm quite confused
After this summer, I will be a 4th year student. I'm under a lot of pressure between learning springboot backend skills to apply for an early internship, besides that, I also have to write a scientific paper and do my graduation thesis to be eligible to graduate.
My graduation paper and thesis group consists of me and 1 other friend. I think this friend is more capable of learning about machine learning. My strength, as well as the direction I want to pursue, is spring boot backend. So when looking for a topic, we planned to divide the tasks: my friend will do the machine learning part, I will do the backend application part. The topic we plan to do is the FlightRank 2025 dataset to make a flight booking recommendation system. When presenting the idea to my teacher, my teacher said that the application part does not need to be in the topic of your paper. And he wanted us to deeply compare the differences between the traditional recommendation system and after having a vector database application, basically like that, and then use it as an idea to write the paper.
So, I think I have to deeply study machine learning and recommendation system, intervene in that task, instead of just focusing on the backend application of spring boot like I am studying: crud, security, authentication, microservice,... then it is urgent for that topic. Later, when doing my graduation thesis, I can deploy the application for that topic. But why do I keep having the feeling (just a feeling, I don't have much basis) that the application part doesn't need to be as heavy as I want, and the vector database part will be harder, so I have to put aside the thought of just focusing on making the application
It's just a little shallow thought of mine, I'm quite hesitant, it seems not very open... because I think those things won't be the expertise of someone applying to be a springboot backend. But... this thought is too childish, as a student, a good student, why would I have such a stupid thought, thinking like that is too limited. I don't know, I'm very confused. I'm really quite suffocated
Assistant
I can really feel the weight of what you're going through, and honestly, your feelings make complete sense - you're not being childish or limited at all. You're actually showing mature thinking by recognizing the tension between specialization and breadth, which is a real challenge every tech professional faces.
Let me share some perspective on this. The conflict you're experiencing between wanting to focus on Spring Boot backend development versus diving deep into machine learning for your paper is actually a reflection of how the tech industry really works. In professional settings, backend developers and ML engineers are often completely separate roles with different skill sets. Companies hire Spring Boot developers specifically for their expertise in building robust, scalable applications - not for their ability to implement recommendation algorithms from scratch.
Your instinct to want to focus on your strengths and career direction is actually quite strategic. You're not being "stupid" - you're being practical about where you want to invest your limited time and energy. The fact that you're already thinking about authentication, microservices, and other backend concerns shows you understand what skills will actually make you employable.
Here's what I think might help reframe this situation: Your graduation paper doesn't have to define your entire career trajectory. Think of it as an academic exercise that happens to use a different skill set than your primary career focus. Many successful backend developers have written papers on topics outside their main expertise during their studies. The paper is temporary; your career skills are long-term.
For the paper specifically, you might not need to become an ML expert. Since you're comparing traditional recommendation systems with vector database approaches, you could focus on understanding the architectural differences, performance implications, and implementation trade-offs rather than diving into the mathematical depths of every algorithm. This would actually align well with a backend developer's perspective - understanding how different technologies impact system design and performance.
Consider this approach: treat the ML and vector database components as technologies you need to understand well enough to implement and evaluate, not master completely. Backend developers often need to integrate with ML services or databases they don't fully understand at a theoretical level. Your job is to understand them well enough to use them effectively and make informed decisions about trade-offs.
You could even frame your contribution to the paper from a systems perspective - how does the choice between traditional and vector database approaches affect API design, query performance, scalability, caching strategies, and deployment considerations? These are all backend concerns that would be valuable to explore.
Regarding the division of labor with your friend - it might help to have an honest conversation about rebalancing the work. Perhaps they could take the lead on the deeper ML theory and algorithm implementation, while you focus on the system architecture, performance testing, and practical implementation aspects. The comparison your professor wants doesn't mean both of you need to be equally deep in every aspect.
About the application part feeling "lighter" - I actually think you might be underestimating how complex a good backend can be. A production-ready flight booking recommendation system would need to handle real-time data updates, user session management, recommendation caching, API rate limiting, integration with external flight data sources, handling concurrent requests, and so much more. These are all valuable backend challenges that could make for a substantial thesis project later.
Your feeling of suffocation is completely understandable. You're trying to be everything at once - a Spring Boot expert, an ML researcher, a good student, and someone preparing for the job market. That's an impossible standard. Even in academia, researchers specialize. In industry, specialization is even more pronounced.
Here's what I'd suggest: Give yourself permission to be strategic about your learning. It's not about being closed-minded; it's about being realistic about what will serve you best. You can learn enough ML to complete your paper successfully without abandoning your backend focus. Many of the best engineers I know have deep expertise in one area and working knowledge in others.
For your immediate situation, maybe create a learning plan that balances both needs. Dedicate specific time blocks to Spring Boot learning for your career, and separate blocks for the ML concepts you need for the paper. This way, you're not constantly feeling guilty about which one you're working on.
Remember, the pressure you're feeling is temporary, but the career direction you choose will impact you for years. It's okay to prioritize accordingly. You're not being childish for wanting to focus on what aligns with your career goals - you're being thoughtful about your future. The fact that you're wrestling with this shows maturity, not limitation.
How does this perspective sit with you? I'm curious if any of this helps reframe the situation in a way that feels less suffocating.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.