KoNoSaiPL/ai-ml-code-interviewer
🧠 Practice coding and quiz yourself on machine learning and deep learning topics with this interactive app designed for interview preparation.
What's novel
🧠 Practice coding and quiz yourself on machine learning and deep learning topics with this interactive app designed for interview preparation.
Code Analysis
14 files read · 5 roundsA Streamlit-based ML/DL interview preparation app that uses LLMs to generate coding exercises with adjustable difficulty, multiple-choice quizzes, code execution, and AI-generated explanations across 30 ML/DL topics.
Strengths
Clean modular architecture with well-separated concerns (LLM service, coding, quiz, history, settings), multi-provider LLM abstraction with graceful fallback to raw HTTP requests, and a thoughtful code execution sandbox with restricted builtins. The dev tooling setup (bandit, detect-secrets, mypy, sphinx) shows professional discipline.
Weaknesses
Zero test coverage despite a thorough dev-requirements setup, the code executor's 'sandbox' is a thin wrapper around exec() that provides minimal real security, and the core value proposition (code generation, quiz generation) is entirely delegated to LLM prompts with no local algorithmic logic or verification of LLM output correctness.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
14
Contributors
3
Files
26
Active weeks
3
Repository
Language
Python
Stars
1
Forks
0
License
—