Recruitron
An AI-powered mock interview simulator offering autonomous question generation, voice-based interactions, and candidate rankings.
The Problem Context
Students and job seekers struggle to practice mock interviews in realistic environments that adapt dynamically to their unique career backgrounds.
The Proposed Solution
Designed a conversational mock interviewer. Leveraging Next.js API routes, candidate resumes are parsed and piped into the Gemini API to formulate customized behavior-driven questions. Speech recognition captures spoken feedback, returning structured analytical scorecards.
System Architecture & Patterns
Engineering Challenges
Ensuring low-latency voice capture and matching the speed of speech recognition with AI question stream chunks.
Key Features
- •Dynamic questionnaire generation tailored to uploaded candidate resume and role definition
- •Voice-to-text response capturing and speech synthesis for immersive interviewer voices
- •Automated evaluation algorithms assessing technical, communication, and analytical skills
Impact
Featured as a major capstone project, used by 120+ peer college students to prepare for upcoming placements.
Lessons Learned
UX needs to account for user silence. Designing fail-safe recovery steps for speech recognition dropouts makes browser speech applications far more reliable.