Sakib Hasan Alif

I’m a Computer Science & Engineering student building practical software and research-focused AI systems. My work centers on CNN-based intrusion detection—where I aim for measurable gains using rigorous evaluation, not just headline accuracy.

In my course research project, I replicated and enhanced a published CNN architecture for network intrusion detection (NSL-KDD). I introduced BatchNorm, a 4th Conv stage, class-weighted loss, gradient clipping, and a dual-optimizer comparison, improving U2R detection rate from 8.96% to 34.33% (3.83×).

I also develop production-style features in my full-stack project “BringIT”—a campus peer-to-peer delivery platform. I implemented JWT auth, real-time WebSocket hooks, leaderboard gamification, and deployed it on Vercel with Supabase Realtime for live order tracking and in-app messaging.

On the engineering side, I’ve designed a 5-stage GitHub Actions CI/CD pipeline (Lint → E2E → Docker Build/Push → Trivy Security Scan → Status Check). I’m driven by clean SDLC practices, strong database fundamentals, and shipping systems that are secure and easy to iterate on.