B.E.W.A.R.E

The project

B.E.W.A.R.E was a 7-person interdisciplinary capstone combining a machine-learning fraud detection model with a web app and a public awareness campaign, tackling India's rise in fake job postings. Scam complaints rose 250% between 2021 and 2023, with average victim losses of ₹15,000 to ₹50,000. The ML side of the project reached strong detection performance (AUC-ROC 0.9983, F1-score 0.98 on the held-out test set). The model, the app, and the campaign are all documented in the full project report ↓, and the final presentation deck ↓ is the shorter walkthrough.

My role

I owned the go-to-market: social strategy, content, and brand identity for the Instagram launch. I built the brand from scratch: a "SCAM stamp" identity in neon yellow, charcoal, and white, paired with a meme-literate, Gen-Z tone to make a serious topic feel approachable instead of alarmist. We ran it hyperlocal at FLAME first: educational carousels mixed with shareable, relatable formats, an on-campus activation booth, and a data-sharing briefing with FLAME's Career Services Office to fold real reporting infrastructure into the campaign.

Results

• 35,400+ views and 13,779 accounts reached in 30 days, starting from zero followers
• 92% of views came from non-followers, so the content traveled on its own
• Top reel hit 17,300+ views; profile visits up 6,324% over the campaign window
• Reels drove 76.5% of reach, followed by Posts (18%) and Stories (5.5%)