This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend
What I Built
I built Interview Prep Buddy, a privacy‑first web app that helps my friend prepare for job interviews. The tool lets them upload their resume and paste a job description, then generates commonly asked interview questions tailored to their skills and experience. It also highlights strengths and gaps in the resume for that specific role, so they can prepare compensating answers.
The app goes further with a mock interview mode: an AI voice asks questions, the user responds via speech, and the system scores their answers, identifies gaps, and adds follow‑up questions. It’s a supportive, interactive way to practice interviews without risking personal data.
Demo
Code
Interview Prep Buddy
Interview Prep Buddy matches a resume to a job description, creates a role-aware question bank, and scores spoken or typed practice answers. Question generation runs in Ollama on each visitor's own device; the app does not use a hosted inference service.
Requirements
- Node.js (only needed to run the local static development server)
- Ollama installed on the device running the browser
- The
llama3.1:8bmodel
Download the model in a terminal:
ollama pull llama3.1:8b
Keep Ollama running while using the app. The app connects from the browser directly to http://localhost:11434; resume and job-description text are not sent to this project's web server.
Run the hosted website with local Ollama
-
Install Ollama on the device running your browser and download the model:
ollama pull llama3.1:8b
-
On Windows, allow the hosted website's exact origin in Ollama. In PowerShell, run:
setx OLLAMA_ORIGINS "https://www.interview-prep.online"If
OLLAMA_ORIGINSalready has values, add…
How I Built It
- Open‑Source AI: Powered by Llama running locally via Ollama.
- Frontend: Minimalist single‑page web app (HTML, CSS, lightweight JS).
- Speech Integration: Browser‑based speech recognition and synthesis for interactive interviews.
- Evaluation: AI + rule‑based scoring for clarity, technical accuracy, and confidence.
The USP is that resumes and job descriptions are processed entirely on the user’s device. Nothing is sent to external servers, ensuring complete privacy.
Why Does Open Innovation Matter?
Open innovation made this project possible. Running open‑weight models locally means users don’t depend on closed APIs or external services. They keep control of their data, while still benefiting from advanced AI capabilities. This freedom to adapt, extend, and run inference locally is exactly why open‑source AI matters — it empowers individuals to build tools that are both useful and trustworthy.
Prize Categories
- Best Use of GitHub Copilot

Top comments (1)
This tool is very helpful as it gives me a better understanding of my strengths and helps me identify the areas where I can improve. It also provides valuable insights into whether my skills and experience are a good fit for the job role. Overall, it is a useful tool for understanding my career suitability and preparing myself for the right opportunities.