fake job detector

• 

  AI-powered job scam analyzer in action—upload or paste a job offer to detect red flags instantly

• 

  Comprehensive scam analysis result—risk score, red flags, and LLM insights all in one view

• 

  Final scam verdict with detailed red flags, company info extraction, and safety tips for job seekers

🧠 Project Story: Fake Job Detector 🚀 Inspiration With the rise of remote work and online job applications, fake job scams have become more sophisticated and widespread. Many people—especially freshers and students—fall victim to job scams that look convincingly real. We wanted to create something meaningful that could protect people from these fraudulent offers and empower them with the tools to verify job authenticity in seconds.

🔍 What It Does Fake Job Detector is an AI-powered tool that analyzes job offers and descriptions to determine if they are legitimate or scams. It checks for:

Suspicious email addresses or phone numbers

Company legitimacy (via LinkedIn and official website checks)

Unusual language, salary claims, or red flags using LLM-based analysis

A scam risk score based on multiple parameters

Users can paste job descriptions or upload offer letters (PDF/DOCX), and the system provides an instant report with a scam score, red flags, and AI insights.

🏗️ How We Built It Frontend: Built using Streamlit for a lightweight, interactive UI

Backend: Python-based logic with modular agents for scam checks, LLM insights, and company legitimacy checks

NLP/LLM: We integrated a reasoning-based LLM agent for understanding the tone and trustworthiness of job posts

Verification Layer: A mini-database of known scam contacts (emails and phone numbers) for real-time checking

File Handling: Support for both PDF and DOCX uploads and automatic text extraction

⚔️ Challenges We Faced Designing a robust risk scoring algorithm that weighs red flags and LLM reasoning logically

Making the company legitimacy checks reliable, especially without access to premium APIs

Handling various file formats and edge cases in poorly structured offer letters

Ensuring the LLM analysis avoids hallucination and remains relevant to scam detection

🎓 What We Learned How to build a modular agent-based architecture using Python

Using LLMs not just for content generation but for intelligent reasoning

How to balance user experience with technical depth using Streamlit

Importance of cybersecurity and awareness in the job hunting space

🌟 What's Next? Add scam report feedback loop for users to improve our model

Deploy as a web app and mobile-friendly version

Integrate real-time scam API checks and public warning alerts

Built With

• llm
• natural-language-processing
• python
• streamlit