-
-
Home/main screen
-
Quiz/question screen
-
Result screen
-
Result screen
-
Result/Score screen
Inspiration
Preparing for interviews can feel repetitive and directionless — most people just Google generic questions and guess whether their answers are good enough. I wanted to build something that gives real practice with real feedback, not just a static list of questions.
What it does
AI Interview Prep Quiz lets users select a field (currently Python Developer and HR) and a difficulty level (Easy or Medium), then answer a set of interview questions. After completing the quiz, the app compares each answer against a sample answer and shows a match score, along with an overall score for the session.
The app has two modes:
Question Bank Mode — pulls from a curated, hand-written set of questions and answers AI Mode — generates completely fresh interview questions in real time using Google's Gemini API, so practice never feels repetitive
If the AI call ever fails, the app automatically falls back to the static question bank, so the experience never breaks.
How I built it
Built entirely in Python using Tkinter for the GUI. Questions are stored in a structured JSON file for the static mode, and the AI mode calls Google's Gemini API with a prompt requesting freshly generated questions in a matching JSON format. Scoring is done through keyword-overlap matching between the user's answer and the sample answer.
Challenges I ran into
Setting up and debugging the Gemini API integration was the biggest challenge — I hit authentication issues with a newer API key format, deprecated library versions, and had to problem-solve through several rounds of trial and error to get a stable connection. I also learned the hard way about keeping API keys out of version control and had to clean up my Git history after accidentally exposing a key.
What wI learned
This was my first time integrating a real AI API into a working application, and my first hackathon overall. I learned a lot about environment variables, API authentication, Git security practices, and building a complete, structured Python application from scratch — from planning, to building screen by screen, to shipping something that actually works end-to-end.
What's next for AI Interview Prep Quiz
. Speech-to-text analysis so users can practice spoken answers . Filler word and speaking pace detection . Additional fields and a "Hard" difficulty tier . Basic engagement tracking via webcam
