Inspiration
Interview preparation can be stressful for students and job seekers. Most preparation tools provide generic questions, but real interviews are different: interviewers ask questions based on your resume, target role, previous answers, and performance.
We wanted to create an experience that feels closer to facing a real interview panel, where different AI interviewers evaluate different aspects of a candidate and adapt the interview based on their responses.
What it does
HireMind (AI Interview Panel) is a multi-agent AI interview platform that helps students and job seekers practice realistic job interviews.
Users can:
- Upload their resume and extract skills, education, projects, and experience
- Select a target job role
- See how their resume matches the selected role
- Practice with multiple AI interviewers
- Answer HR, technical, and behavioral questions
- Practice coding questions
- Use voice or typing to answer questions
- Receive feedback on their answers
- Practice with timed questions
- Track their interview performance over time
- View an overall readiness score
The interview panel consists of four specialized AI agents:
- Manager Agent — manages the interview and decides which interviewer should ask the next question
- HR Agent — evaluates motivation, culture fit, salary expectations, and other HR topics
- Technical Agent — asks role-specific technical questions based on the candidate's skills and projects
- Behavioural Agent — evaluates teamwork, leadership, conflict resolution, and decision-making
The system adapts the interview experience based on the candidate's responses and provides actionable feedback after each answer.
How we built it
HireMind uses a modern web application architecture with a frontend and AI-powered backend.
The frontend provides the interview interface, resume upload, job-role selection, question flow, coding interface, voice interaction, feedback, and progress tracking.
The backend handles resume processing, AI interactions, interview logic, agent coordination, and candidate evaluation.
The system follows a multi-agent architecture where specialized AI agents handle different parts of the interview.
The main workflow is:
- The candidate uploads a resume.
- The system extracts relevant skills, education, projects, and experience.
- The candidate selects a target job role.
- The system compares the resume with the selected role.
- The Manager Agent controls the interview flow.
- HR, Technical, and Behavioural Agents conduct specialized interview rounds.
- The system evaluates the candidate's responses.
- The candidate receives feedback, scores, weak areas, and a readiness assessment.
Technologies used include React, Vite, JavaScript, Python, FastAPI, Axios, PDF processing, and an LLM API.
Challenges we ran into
One of the biggest challenges was designing a multi-agent interview system where every agent has a clear responsibility while still creating one continuous interview experience.
We also needed to connect resume information with the selected job role so that technical questions could be relevant to the candidate's actual skills and projects.
Another challenge was making the interview feel realistic while keeping the interface simple enough for students and job seekers to use without technical knowledge.
Supporting different interaction modes, including typing, voice responses, coding questions, and timed questions, also required careful coordination between the user interface and the AI interview workflow.
Accomplishments that we're proud of
We are proud of creating a working AI interview platform that goes beyond a simple chatbot.
The project combines:
- Resume analysis
- Role matching
- Multi-agent interviews
- HR evaluation
- Technical evaluation
- Behavioral evaluation
- Coding practice
- Voice interviews
- Timed questions
- Answer feedback
- Progress tracking
- Readiness scoring
into a single interview preparation experience.
The Manager Agent allows the different interview agents to work together instead of treating every question as an isolated chatbot interaction.
What we learned
Building HireMind taught us how multiple specialized AI agents can work together to solve a larger problem.
We learned how to connect an AI-powered backend with a modern React frontend, process resume PDFs, use candidate information to generate relevant questions, and design an interactive AI workflow.
We also learned that creating a useful AI application is not only about generating responses. The system needs clear agent responsibilities, structured workflows, useful evaluation criteria, and a simple user experience.
What's next for HireMind
We plan to continue improving the platform with:
- More advanced voice and speech analysis
- Deeper communication analysis
- Video interview analysis
- More job roles and industries
- Advanced coding assessments
- Interview history and analytics
- More personalized interview difficulty
- Improved readiness predictions
- Personalized learning plans based on weak areas
Built With
- api
- axios
- fastapi
- javascript
- llm
- multi-agent
- processing
- python
- react
- vite
Log in or sign up for Devpost to join the conversation.