StdOut

StdOut lets engineers practice real technical interviews with live AI interviewers, coding challenges, and in depth feedback while replicating the pressure and experience of actual interviews.

Comment

Inspiration

As penultimate year Software Engineering students who are chasing internships, we found ourselves unprepared for one critical step, the technical interviews. They began to feel unfamiliar, high pressure, and hard to practice realistically. We decided to build this platform to recreate the experience so students can practice, gain confidence, and walk into interviews ready.

What it does

Our web app is an interview preparation platform that simulates a realistic technical interview environment. Users are presented with both coding and conceptual questions while an AI interviewer listens to their verbal reasoning and asks follow up questions to their written code and reasoning in real time through a live transcription. After completing a question, the system evaluates both the written solution and the spoken explanation, generating a structured review with feedback on correctness, clarity and problem solving approach.

Our Community

We are building for the community of aspiring software engineers, especially students and early career developers who often lack access to the realistic technical interview practice. Many rely on fragmented resources or informal prep, making it hard to build confidence in real interview settings. Our goal is to create a supportive and accessible platform that empowers this community to practice together, improve their skills, and succeed in the hiring process.

How we built it

We built the application using a full stack web architecture. The frontend is developed with React to provide an interactive and responsive user experience. The backend uses Node.js and Express to handle API requests, session management and evaluation workflows. MongoDB was used to create a quick and ready to use database based on schemas. We utilised OpenAI models for live text transcription, real time interviewer and answer feedback. For deployment, Vercel, Render and MongoDB Atlas were all leveraged to create a fully functional and deployed application.

Challenges we ran into

One of the main challenges we encountered integrating and connecting our database. We attempted to do this on the first day at around 1am and due to our sleep deprivation we struggled to get it to work. Furthermore, we also faced difficulties creating a realistic AI interviewer that could simulate authentic interview reactions. Particularly, we had issues with getting the interviewer to talk about the problem with the end user without revealing the answer. Our final challenge was finding and configuring a tech stack for deployment. Our final decisions were based on cost effective and simple deployment ranks, however this still led to some challenges as it was a completely new experience for the 3 of us. As a team we grunted through the stages of changing all APIs from local host to our respective frontend, backend, and deployment, while running into what felt like an endless stream of errors.

Accomplishments that we're proud of

We are particularly proud of the clean and intuitive user interface we were able to design. We believe we have made the application both engaging and easy to use. Additionally, the integration of AI throughout the platform is a standout achievement. This enabled us to implement some really cool features such as real time interaction, intelligent feedback and a more immersive interview experience.

What we learned

Starting with the technical side, our team comprises members who are familiar with Data Analytics and System design, not App development. So embarking on a full stack project was very new and unfamiliar for us. We had done some small projects here and there with frontend, backend and a database but have never made one to this scale. We also have never deployed an application before. Surprisingly though, our biggest learning was how to work in a small team under immense time pressure. We learnt how to discuss ideas, delegate tasks, and code on the same project simultaneously. Throughout this journey team members began taking lead at different times and we effectively worked together to create our deployed app.

What's next for StdOut

In the future, we plan to expand StdOut by incorporating system design questions to support more advanced technical interview preparation. We also aim to make the AI interviewer more personable and adaptive, allowing it to adjust its tone, difficulty, and follow up style based on the user’s performance and experience level. Beyond this, we want to introduce personalised learning paths that track user progress over time and target weak areas with tailored questions. Additional improvements include enhanced speech analysis and support for multi round mock interviews that better simulate the hiring process. Ultimately, our goal is to create a fully immersive end to end interview preparation platform that evolves with each user.

Note: The demo video failed to record the voice of the interviewer. If you wish to view this function try the app yourself using the link provided.

Built With

Share this project:

Updates