MockInterviewerAI

Inspiration

While preparing for Software Engineering interviews, I noticed that there is no go-to tool for mock interview practice. Mock interviewing on online platforms is often expensive and isn't always a frequently and easily available option for many. This is one AI mock interviewer tool that would help you with real-time mock interviews simulated like how they are conducted at most of the Big Tech companies, be it Coding Round, Machine Coding/LLD round, System Design/HLD Round and Behavioral Round and that too across different levels.

What it does

This tool acts like a real technical interviewer and can see what code you write or what architecture diagram you draw, listen to your solution approach, check for plagairism or tab switch during the interview, review your code and provide detailed downloadable feedback, where all interviews are timed.

How I built it

This project has mostly been vibe coded in Google AI Studio with the following Technology Stack: Frontend: React 19, TypeScript, Vite, Tailwind CSS, Framer Motion Editor & Canvas: @monaco-editor/react, react-konva AI Engine: @google/genai (Gemini 2.5 Flash Native Audio Preview for Live Voice) Gemini 3 Flash Preview (for dynamic question generation and feedback reports) Utilities: html2canvas (screen capture), jspdf (report generation) Hosting: Firebase Hosting on Google Cloud Platform (GCP).

Challenges I ran into

Initially I faced some problem or errors with Firebase Deployment but I troubleshooted the situation successfully. Later, I ran into some error with the referrer header which was solved by removing the Application Restrictions, and by setting the API restrictions in Google Cloud Console.

What I learned

From a technical standpoint, I learned how to manage complex React state alongside real-time WebSocket streams and how to effectively prompt multimodal models. More broadly, I learnt that, AI can be a powerful and tremendously helpful tool for practicing mock interviews effectively, frequently and cost effectively.

What's next for MockInterviewerAI

In the future versions of MockInterviewerAI, I would like the system to remember what problems I have already solved in previous mock interviews and accordingly ask me questions I haven't solved on the platform previously. This would require the system to be stateful and would utilize a database. I also plan to add specific "Company Personas" to simulate the unique interview styles of different Big Tech companies.

Built With

• firebase
• framermotion
• gemini2.5flashnativeaudiopreview
• gemini3
• react
• tailwindcss
• typescript