The major problem that Code Companion aims to tackle is to eliminate the tendency to base our own logic off another person's logic. We immediately turn to the provided code logic or from Youtube videos as we find developing our solution as cumbersome & time-consuming. This interface works purely on improving our logic to finally help us arrive at a logic that covers the required time & space complexities by adding on suggestions to the idea that we fed into it. It makes this concept of competitive coding more fun & less tiresome as we don't have to feel stressed out about our logic not reaching somewhere as the platform constantly tries to get our solution to its pinnacle efficiency. We reap the benefits of a combine study or the presence of a person critisizing our opinions constructively and providing add-ons to it.Such a system helps us depend more on our own logic and finally crafts us into a reliable developer.
This need not stop at just a LeetCode edition. It can be extended to any self-learning activity using dedicated training of an AI model which is more effective than OpenAI. Such a solution can largely change the scenario in the technological world as more unique logics come out as it is rooted from our own logic.
The main challenge I ran into was failing to complete this project of mine.The project is not yet complete as it has not formulated into a conversation-level platform. Various security issues where faced while using OpenAI, which I oveercame by using OpenAI using only python functions due to security reasons. This was the main reason for failing to complete the project as creating the desired coversation UI is much easier in javascript than in python as dynamic text generation is much easier in javascript.
Dynamic URL routing in Django for each question was also quite a task but was sorted out through learing more about Django.Creating an appropriate UI for this project was yet another hurdle which consumed a lot of time.
The inability to retai the memory of previous prompts & answers was yet another obstacle as it would prevent the generation of a conversation-like platform.This was solved by allocating an empty variable for each question which is used to store the previous result and query & hence can retain the memory even after days without usage,
Tracks Applied (1)
Technologies used
Discussion