Inspiration
Preparing for interviews is stressful because candidates rarely get the chance to practice in a realistic setting. Most existing tools rely on static question lists or generic advice. We wanted to create an experience that feels like a real interview where users speak, respond to meaningful follow-up questions grounded in their own resume and the job they’re applying for.
What it does
Interview Preparation Coach is a conversational interview-practice web app. Users upload a job description and their resume, then practice interviews by speaking or typing their answers. The system transcribes spoken responses, maintains conversational context, and asks adaptive follow-up questions tailored to the role and the candidate’s background.
How we built it
The app is built using the following architecture:
- Frontend: Streamlit for rapid interactive prototyping
- Speech-to-Text: Whisper for accurate, real-time transcription
- Language Model: Gemini for interviewer-style question generation and follow-ups
- Context Grounding: Resume PDFs are converted to text and combined with the job description
Accomplishments that we're proud of
- Built a fully conversational interview experience with both voice and text input
- Successfully grounded AI responses in user-provided resumes and job descriptions
- Designed a clean, modular architecture that separates UI, transcription, and LLM logic
- Delivered a responsive, demo-ready system despite platform-specific challenges
What we learned
We learned that realistic AI experiences require more than strong models. They require thoughtful system design. Using specialized tools for speech and conversation dramatically improved reliability. We also gained practical experience handling multimodal pipelines and debugging real-world dependency issues.
What's next for Interview Preparation Coach
Future work includes adding structured feedback and scoring, interviewer personas (technical, behavioral, or friendly), session saving for progress tracking, and real-time confidence or pacing analysis to further enhance the interview practice experience.
Log in or sign up for Devpost to join the conversation.