Genuire: An AI-Powered Job Scam Detection Platform

Souvik Das — Mon, 05 Oct 2026 04:32:55 +0000

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend

What I Built

Genuire is an advanced, real-time AI/ML platform designed to protect job seekers from fraudulent employment opportunities. I built this for a close friend who is actively navigating the difficult job market and recently encountered a highly sophisticated employment scam.

Genuire solves this problem by actively auditing job descriptions, verifying URLs and corporate registrations, and analyzing recruiter contact channels. It acts as a shield, calculating a comprehensive Trust Score (0-100%) using a 5-Pillar Trust Engine before the applicant submits any sensitive data, ensuring they don't fall victim to phishing or identity theft.

Demo

genuire.onrender.com

Code

rashmi-crypto / Genuire

Genuire: Fake Job Posting Detection Platform

Genuire is an advanced, real-time AI/ML platform designed to protect job seekers from fraudulent employment opportunities. By actively auditing job descriptions, verifying URLs and corporate registrations, and analyzing recruiter contact channels, Genuire calculates a comprehensive Trust Score (0-100%) before applicants submit their sensitive data.

✨ Key Features

Real-Time Threat Detection: Instantaneous analysis of job postings to identify potential scams.
Multimodal Gemma Analysis: Uses Google's gemma-4-31b-it vision models to scan uploaded screenshots of job postings for visual and textual fraud indicators.
Smart URL Scraping: Automatically fetches and populates job details (title, company, description) from major job boards for seamless analysis.
5-Pillar Trust Engine: Employs a multi-faceted verification pipeline incorporating machine learning and heuristic risk metrics.
Premium Visual Dashboard: A sleek, responsive user interface featuring a dynamic horizontal pipeline tracker, automatic Light Mode, and detailed trust telemetry.

🛠️ System Architecture

Genuire…

View on GitHub

How I Built It

Genuire is built around a hybrid AI architecture. It leverages Google's open-weights gemma-4-31b-it vision model to power its multimodal analysis. When a user uploads a screenshot of a suspicious job posting, Gemma scans it for visual and textual fraud indicators.

The open-source AI is complemented by a high-performance scikit-learn machine learning pipeline. It uses an SGDClassifier for NLP text analysis and a RandomForestClassifier for tabular risk data (like location threat ratios and character lengths). The application is glued together using a Flask backend and a responsive Vanilla JS/CSS frontend featuring a dynamic 7-stage verification pipeline tracker.

Why Does Open Innovation Matter?

Open innovation is crucial for security and fraud-prevention tools. By utilizing open-weights models like Gemma and open-source libraries like scikit-learn, developers can completely scrutinize the model's biases and fine-tune its capabilities specifically for threat detection without relying on a black box.

A closed API would make it difficult to transparently explain why a particular job posting was flagged or approved. Open innovation made it possible to build a transparent "Trust Engine" where the user can see exactly which of the 5 pillars passed or failed, giving them the confidence and context they need to make safe career decisions.

My Agent Session

For this project, I pair-programmed extensively with my AI coding agent (Google Gemini/Antigravity). The agent was instrumental in helping me architect the 5-Pillar Trust Engine, fine-tune the scikit-learn models for better accuracy, and rapidly prototype the Flask backend. We also collaborated closely to polish the CSS for the dynamic horizontal pipeline tracker on the frontend, turning a complex ML backend into a highly premium, user-friendly dashboard!

Prize Categories

Gemma - Best Use of Gemma: Genuire integrates Google's open-weight gemma-4-31b-it model to conduct advanced multimodal image analysis on job posting screenshots, extracting visual and textual fraud indicators.
Render - Best Use of Render: The entire Genuire Flask application and AI runtime is configured for seamless deployment and hosting on Render, as defined by our render.yaml configuration.
Entire - Best Use of Entire: As detailed in the 'My Agent Session' section above, I utilized an AI coding agent to architect the core Trust Engine and build out the frontend styling, sharing that session directly in this write-up.

Team Members

A huge thanks to my amazing teammates who helped bring Genuire!

@rashmipyne

DEV Community: Souvik Das