Inspiration
Interviews can be stressful, especially for students who don't have access to continuous interview practice or personalized feedback. We wanted to build an AI-powered platform that allows students to practice interviews, identify weaknesses, and improve their confidence in one place. The idea was to make interview preparation feel more interactive than simply reading questions or watching tutorials.
What it does
AI Interviewer is an interactive interview preparation platform that simulates real interview experiences and provides personalized feedback.
Interactive AI Interview – Users practice real-world technical and behavioral questions through a dynamic interview interface. Mock Test & Feedback – Users can take mock interviews/tests and receive feedback on their performance. Sentiment Analysis – Uses TextBlob to analyze responses and provide insights into tone and confidence. Filler Word Detection – Detects words such as “um,” “uh,” and “like” to help users communicate more professionally. Technical Quiz Suite – Topic-specific MCQs covering Web Development, Backend, APIs, and AI/DS/ML. Readiness Score – Provides a percentage score and classifies users as Beginner, Intermediate, or Ready. Portfolio Intelligence – Scans projects and identifies 20+ technical skills and professional action verbs. Strategic Feedback – Provides suggestions to strengthen the user's interview performance and portfolio. Glassmorphic UI – A modern, responsive interface designed for both desktop and mobile.
How we built it
We built the application using a lightweight full-stack architecture:
- Frontend: HTML5, CSS3, JavaScript
- Backend: Python + Flask
- AI/NLP: TextBlob for sentiment and language analysis
- Deployment: Render
- Production Server: Gunicorn
The Flask backend handles the interview logic, quiz evaluation, sentiment analysis, filler-word detection, and portfolio analysis, while JavaScript provides an interactive frontend experience.
Challenges we ran into
One of our biggest challenges was turning traditional interview preparation into an interactive experience rather than a simple question-and-answer system.
We also had to figure out how to provide meaningful feedback from text responses. Detecting sentiment and filler words while keeping the feedback understandable and useful required careful processing.
Another challenge was designing a UI that looked modern while remaining responsive and easy to navigate across different screen sizes. Deploying the Flask application with Gunicorn and Render also required configuring the application for production.
Accomplishments that we're proud of
We are proud of creating an all-in-one interview preparation platform rather than focusing on only one feature.
The project combines:
Mock Interviews + Technical Quizzes + AI/NLP Feedback + Filler Word Detection + Portfolio Analysis + Readiness Scoring
We are especially proud that the application is deployed and accessible online, allowing users to actually interact with the system instead of it being only a local prototype.
What we learned
Through this project, we learned how to combine frontend development, backend development, and AI/NLP techniques into a complete application.
We gained practical experience with Flask, TextBlob, JavaScript, API-style backend logic, deployment using Gunicorn and Render, and designing responsive user interfaces.
Most importantly, we learned that building an AI-powered application is not only about implementing AI — the feedback needs to be understandable, actionable, and useful to the user.
What's next for AI Interviewer
We want to make AI Interviewer more intelligent and personalized.
Future improvements include:
Voice-based interviews with speech recognition Body-language and facial-expression analysis More advanced AI-generated interview questions Personalized interview difficulty based on performance User progress tracking and performance history Resume-based interview generation More detailed AI feedback after every answer Leaderboards and interview readiness tracking
- 🔗 Integration with real job descriptions to generate role-specific interviews
Our ultimate goal is to turn AI Interviewer into a personal AI interview coach that continuously adapts to each user's strengths and weaknesses.
Built With
- ai
- css3
- flask
- git
- github
- gunicorn
- html5
- javascript
- machine-learning
- natural-language-processing
- python
- render
- restapi
- sentimentanalysis
- textblob
Log in or sign up for Devpost to join the conversation.