Practice resume-specific interviews with a local Qwen model
At a glance
FriendPrep AI turns a candidate's resume and target job description into mock questions and feedback using local inference. This workflow is listed under Job Search, Interview Preparation, Local AI.
- Agents / tools
- FriendPrep AI, Qwen3, Ollama
SourceBuilding a tool for these workflows? See weekly rates.
What they did
1) Have Git, Python, Node.js and Ollama. Clone https://github.com/Stuti-Singhania/friendprep-ai and run ollama pull qwen3:1.7b-q4_K_M; keep Ollama running. 2) In backend, create and activate a Python virtual environment. Install fastapi, uvicorn, pydantic, pypdf, python-multipart and ollama with pip, then run uvicorn app.main:app --port 8000. In another terminal, run npm install and npm run dev from frontend. These launch commands are derived from the repository's entry points. 3) Open localhost:3000. Upload a text-readable resume PDF under 5 MB and paste a job description of at least 50 characters. Select Generate Interview Questions. 4) Start Mock Interview, answer the first question, and choose Evaluate My Answer. Read the score, strengths, missing concepts and suggested answer; advance through the remaining questions and finish with Review Questions.
Worth knowing
The code uses qwen3:1.7b-q4_K_M despite a different model label in the interface. Scanned resumes without extractable text are rejected. No hiring outcome is reported.
