Inspiration
In today's competitive job market, a well-crafted resume can make all the difference. However, getting expert feedback on a resume is often expensive or time-consuming. We wanted to create a tool that provides instant, AI-powered resume reviews and mock interviews to help job seekers refine their applications and interview skills effortlessly.
What it does
ReviewGPT is an AI-powered resume analyzer and mock interview assistant. Users can:
- Upload their resume (PDF), which is analyzed for strengths and weaknesses.
- Receive a detailed AI-generated review with constructive feedback and suggestions for improvement.
- Choose a tone for their review, whether they want friendly advice, professional critique, or even a humorous roast.
- Engage in a real-time AI-driven mock interview, where the AI interviewer asks relevant questions based on the resume and evaluates user responses.
- Interact via text and voice, using text-to-speech (TTS) and speech recognition for a seamless experience.
How we built it
We developed ReviewGPT using:
- HelpingAI the brain behind the app
- Flask (Python) for the backend API, handling resume text extraction and AI interactions.
- PyMuPDF for accurate PDF text extraction.
- HelpingAI API to generate AI-driven resume feedback and interview questions.
- JavaScript (Frontend) to provide an intuitive UI with file uploads, text display, and speech functionality.
- SpeechRecognition & Web Speech API to enable voice-based interviews.
- Vercel for quick deployment and API hosting.
Challenges we ran into
- Ensuring accurate resume text extraction: Some PDFs had formatting issues, making text extraction tricky.
- Making the AI-generated reviews insightful and structured: Tuning the prompts for different tones and making sure the responses were coherent took several iterations.
- Building a natural AI interview flow: Handling user responses dynamically while ensuring the AI provides relevant follow-up questions was complex.
- Integrating Speech Recognition: Ensuring smooth real-time voice interaction was a technical hurdle.
Accomplishments that we're proud of
- A functional AI-driven resume review system that adapts to user preferences.
- A fully interactive AI mock interview that allows users to practice with real-time speech input.
- Speech recognition integration, making the interview feel more natural.
- A Web app that works smoothly across different devices.
What we learned
- How to extract clean and structured text from PDFs with various formats.
- Web Speech API's limitations and workarounds for smoother voice interactions.
- How to optimize Flask APIs for quick responses when handling AI requests.
What's next for ReviewGPT
- Enhancing AI adaptability: Customizing interview questions based on job roles.
- Scoring system: Providing quantitative feedback on resume quality and interview performance.
- Multi-language support for non-English users.
- Integration with LinkedIn & job platforms for one-click resume evaluation.
- AI-generated resume templates to help users create job-winning resumes instantly.
- Real-time AI coaching to provide suggestions while users answer interview questions.
ReviewGPT is just the beginning of making AI-powered career guidance accessible to everyone!
Log in or sign up for Devpost to join the conversation.