Skip to content
All projects
AI

Fraud Job-Email Detector

Watches your inbox and flags fake job offers with rules plus GPT analysis.

Role
Architect and full-stack developer
Illustration of an inbox with a suspicious job email flagged and a warning shield

What it is

Connect one or more mailboxes and the system checks new email every five minutes, scoring each job offer for fraud on company legitimacy, job description, contact details and salary.

The problem

Job seekers are targeted by convincing fake offers. Rules alone miss new scams, and an LLM alone is slow and costly on every message.

Architecture

Requests flow left to right. Hover a component for its role.

Fraud Job-Email Detector architecturefetchMailboxes: IMAP, Gmail APIMailboxesCelery beat: every 5 minutesCelery beatCelery workersCelery workersRule heuristicsRule heuristicsOpenAI GPT: category scoringOpenAI GPTPostgreSQLPostgreSQLDjango REST APIDjango REST APIReact dashboardReact dashboard
Solid line: synchronous callDashed line: through a queueDashed box: external or infrastructure
  • Mailboxes external, IMAP, Gmail API
  • Celery beat infrastructure, every 5 minutes
  • Celery workers worker
  • Rule heuristics service
  • OpenAI GPT ai, category scoring
  • PostgreSQL data
  • Django REST API service
  • React dashboard client

Key decisions

  • Hybrid detection

    Cheap rule-based heuristics catch known fraud patterns, and GPT analysis scores each category, so the score has explainable parts.

Highlights

  • Multiple mailboxes per user through IMAP or the Gmail API.
  • On-demand rescans and a dashboard of fraud statistics.
  • Email sign-up and Google sign-in.

Stack

Django 5Django REST FrameworkCeleryRedisPostgreSQLOpenAIReact 18ViteMaterial UIRecharts