Fake job post detector

Inspiration

Fake job scams are common on social media, where posters often use the logos of well-known companies to appear real. Many people fall victim to these scams due to attractive phrases like "100% job guarantee" or "no interview required".

What it does

Our tool lets users upload a job poster image. It extracts the text using OCR and checks for common scam phrases. If any suspicious text is found, the system alerts the user.

How we built it

• We used Python and Tesseract OCR (pytesseract) to extract text from the poster.
• Simple rule-based checks were implemented using a list of scam keywords.
• A Streamlit interface allows easy image upload and instant results.

Challenges we ran into

• Extracting clean text from low-quality images was tough.
• Many scam posters use informal or tricky language, so we had to refine our keywords.
• Limited dataset of scam posters meant we had to manually collect and label examples.

Accomplishments that we're proud of

This whole project and concepts.

What we learned

• OCR is powerful but needs good image quality.
• Even simple AI systems can make a meaningful impact if focused well.
• Designing for explainability is important — users liked knowing why a poster was flagged.

What's next for Fake job post detector

• Add logo and contact info mismatch detection
• Include image quality checks (e.g., blurry text)
• Expand dataset and make detection smarter using ML

Built With

• pillow
• python
• streamlit
• tesseract-ocr-(pytesseract)