GEN AI-Powered Socratic Teaching Assistant
Empowering students to master DSA through guided, Socratic questioning powered by GenAI
Created on 2nd October 2024
•
GEN AI-Powered Socratic Teaching Assistant
Empowering students to master DSA through guided, Socratic questioning powered by GenAI
Describe your project
1 . In-Scope:
In-Scope:
This solution utilizes GenAI, specifically the Gemini 1.5 Flash-002 model, to create an AI-powered Socratic teaching assistant tailored for students learning Data Structures and Algorithms (DSA). The assistant covers essential DSA concepts like sorting algorithms, recursion, dynamic programming, and complexity analysis. It guides students through Socratic questioning, helping them develop critical thinking and problem-solving skills. The system also retains conversational memory, adapting to students' answers and adjusting its questioning to enhance engagement and personalized learning.
2 . Out of Scope:
The assistant is restricted to DSA topics only. It doesn’t provide direct coding feedback or support for subjects outside DSA. Real-time code execution, detailed debugging, or advanced personalized tutoring beyond guiding questions is also out of scope.
- uture Opportunities:
The solution could be extended to cover a wider range of computer science subjects, integrate code execution for live feedback, and support personalized learning paths. It could also introduce multilingual capabilities, track student progress, and provide more in-depth assessments or advanced hints to guide students further in their learning journey. The assistant can be expanded into collaborative environments, where multiple students can engage in discussions, enriching peer-to-peer learning experiences.
Challenges we ran into
1 .API Limitations:
Since we used the free API from Google, we faced strict limitations on the number of requests. This often resulted in interruptions during testing, affecting our ability to evaluate the model’s performance effectively.
-
Memory Management Issues:
Despite implementing conversational window memory, the free version of the API didn't allow for robust memory functionality. This limitation hampered the assistant's ability to retain context over multiple interactions, diminishing the user experience. -
Cost of Advanced Features:
While trying to enhance the model's capabilities through Google AI Studio by tuning the model with various input and output, we found that many features required a paid subscription. This made it challenging to optimize our solution without incurring additional costs.
To overcome these challenges, we experimented with different API keys from multiple accounts, which helped us manage usage more effectively. We also made adjustments to our implementation strategy, focusing on optimizing existing features within the constraints of the free API. Ultimately, this process taught us valuable lessons in resource management and adaptability in AI development.
Tracks Applied (1)