Hey, I'm Roshaan — an aspiring cybersecurity engineer energized by digging beneath the surface of systems. I wanted to know how systems communicate, how access is controlled, where weaknesses begin, and what it takes to protect something well. That curiosity led me into computer science; I recently graduated from the University of North Carolina at Charlotte with a B.S. in Computer Science and a concentration in Cybersecurity. I build practical defenses across Linux, cloud security, and network stacks — hardening services, designing IAM controls, and shaping detection that turns signals into action. What keeps me energized now is shipping real projects across Linux, cloud security, network defense, IAM, and IT audit, turning curiosity into hands-on skill. I learn by doing: shipping small, verifiable wins, iterating quickly, and communicating technical findings clearly. I’m especially drawn to work that combines technical depth with structure — investigating signals, hardening systems, and building security with intention from the start.
A few of the systems and technical projects I’ve built to sharpen my skills across cybersecurity, cloud security, Linux, digital forensics, and user-focused product work.
A hands-on cloud security lab focused on AWS Identity and Access Management, least-privilege access, security logging, and threat-detection workflows. I’m building it through progressively implemented scenarios so I can better analyze identity activity and investigate potentially risky events in a cloud environment.
Designed and implemented a Linux firewall architecture using nftables with a stateful default-deny model, controlled SSH/HTTP/HTTPS access, DNAT, container networking, and network segmentation. The project strengthened my understanding of TCP/IP, routing, and testing connectivity with Linux networking utilities.
A reproducible IAM control-testing workbench that turns raw identity evidence into structured, reviewer-ready audit findings. It simulates an IAM audit, retains evidence, applies rule-based severity, and presents the results through an interactive dashboard and audit workpapers.
A construction operations platform built to turn scattered invoices, field notes, and cost updates into structured project intelligence. It supports invoice analysis, CFO-style briefings, and ERP-ready site updates so contractors can see what is approved, what is at risk, and what needs attention.
Four JPEGs, no context. This investigation extracted EXIF data, correlated GPS coordinates, and anchored every step with MD5 and SHA-1 hashing to reconstruct where each file was taken while preserving chain of custody and evidentiary integrity.
An AI-assisted platform concept designed to help users better understand career pathways, skill signals, and growth directions through a more thoughtful and user-friendly experience. The project explores how data and interface design can make career guidance feel smarter and more personalized.
Responsive web application work built for local service businesses using modern browser APIs, maps, location-based workflows, booking functionality, client-side validation, and mobile-first interfaces designed to feel clear, fast, and useful for real visitors.
Participated in professional development activities, technical workshops, networking events, and career-insight sessions that strengthened my communication skills and broadened my exposure to cybersecurity, consulting, and technology careers.
Took part in collaborative technical projects and fast-paced hackathon environments that pushed me to learn quickly, communicate clearly, and turn ideas into working prototypes.
Participated in UNC Charlotte's CCI Hackathon, working in a team-based environment focused on problem-solving, rapid development, and turning technical ideas into a working project.
During the 36-hour AI ATL weekend, I worked as a developer with my team to build a Career Intelligence platform from the ground up. We used AI-powered features to support professional growth and job matching, developed a functional prototype, and presented our solution to industry judges during Demo Day.
Led credit risk analytics within a five-person team, working with 2.26 million LendingClub loan records in Python and Jupyter to identify default patterns, financial exposure, and business-focused recommendations while learning directly from Bank of America professionals.