System Documentation

Technical guide for the JD2Q Interview Intelligence Pipeline.

01 Overview

JD2Q is a specialized pipeline designed to bridge the gap between static job descriptions and dynamic interview preparation. By utilizing the Gemini 2.0 architecture, the system extracts high-value semantic signals to generate balanced question sets covering technical depth and behavioral intelligence.

02 Onboarding Sequence

Step A

Authentication

Initialize access via Google OAuth or OTP-based verification.

Step B

Key Integration

Add your Google AI Studio API credentials to the secure vault.

Step C

Context Submission

Paste raw job description text for initial parsing.

Step D

Synthesis

Generate and refine question matrices based on role seniority.

03 Generation Logic

Core Constraints

  • 01 1,500 word threshold for raw JD input.
  • 02 Structured output ensuring at least 15 balanced questions.
  • 03 Context-aware regeneration (Duplicate suppression).

Note: AI outputs should be reviewed for brand-specific alignment. Use "Generate Answer" for baseline reference signals.

04 Security Protocols

Protocol Area Mechanism
API Key Storage AES-256 Symmetric Encryption (Fernet)
Auth Sessions Supabase JWT + Session-locked Cookies
Rate Limits Redis-based burst protection (50 req/min)

05 Static Architecture

01 Backend: Flask Cluster (Python 3.11)
02 Intelligence: Gemini 2.0 Flash / Pro
03 Database: PostgreSQL (Managed by Supabase)
04 UI System: Tailwind CSS Core + Glassmorphism
05 Deployment: Vercel Pulse V2

End of Protocol • Rev 2.0.4