Interview Copilot

Technical README Description

Interview Live Copilot is Python-based interview assistance system designed to support candidates during live technical and behavioral interviews. The application ingests a candidate resume, extracts structured experience and story data, analyzes the target job description, and retrieves the most relevant evidence from a local vector store. When the interviewer asks a question, the system detects whether the input is a question, classifies its type, and generates a response grounded in the candidate’s actual experience and role requirements.

The backend is built with FastAPI and uses Google Gemini for structured response generation, while a local SQLite database stores interview sessions and transcript metadata. The system also maintains daily and per-session usage tracking to protect free-tier limits and prevent excessive API usage. The frontend is a lightweight desktop/browser interface that surfaces live transcript updates and suggested answers in near real time.

Key components include:

• Resume extraction and parsing via Gemini
• Job analysis and requirement matching
• Retrieval of relevant candidate evidence from a FAISS-backed local index
• Local question detection and classification heuristics
• Conversation memory summarization for follow-up context
• WebSocket-based real-time interview updates
• Desktop wrapper using PyWebView for local execution

Run in a browser

cd backend
..\.venv\Scripts\python.exe -m uvicorn app.main:app --reload

Open http://127.0.0.1:8000/ui/.

Copy backend/.env.example to backend/.env and set GEMINI_API_KEY to enable AI features. On the first run, /api/resume/current extracts the Markdown resume and stores the structured profile locally; later runs reuse the saved candidate profile. The first semantic retrieval run may download the Sentence Transformers model.

Run as a desktop app

..\.venv\Scripts\python.exe backend/desktop.py

PyWebView requires WebView2 on Windows. The browser mode is the simplest development workflow.

Tests

cd backend
..\.venv\Scripts\python.exe -m pytest -q

API highlights

• GET /api/resume/current loads the persistent local resume profile.
• POST /api/job/analyze analyzes a job description.
• POST /api/interview/answer generates an evidence-grounded answer.
• WS /api/interview/ws/{session_uuid} runs the live transcript pipeline.
• DELETE /api/privacy/candidate/{candidate_id} deletes profile and vector data.