♃ CLOUD SECURITY · CODE · AI · DESIGN
Building things. Breaking things.
Securing things.
I'm Sisco Ask, an aspiring Cloud Security Engineer working across code, AI and design. I ship real products, then go find the ways they'd break.
♃ PROFILE
The short version
Sisco Ask — Sholuade AbdulRazaq Oluwatoyin
I'm working toward a career securing cloud infrastructure — AWS, networking, Linux, and the fundamentals underneath it all — while staying hands-on as a builder. Most of what's below started as "let me see if I can actually ship this," not a class assignment.
That means my work spans further than a typical résumé bucket: fintech interfaces, a cryptography workbench, documented security labs, and AI-assisted products. Different surfaces, same instinct — build it properly, then try to take it apart.
♃ TOOLING
What I work with
Cloud & Security
Software Development
AI & Vibecoding
Design
♃ SELECTED WORK
Case files
♃ Security & Cloud
Read the case study
The Problem
Most "learn cryptography" resources are either pure theory or copy-paste snippets with no visibility into what's actually happening to the data. I wanted one tool that ran real WebCrypto operations and showed the real output — so the concepts stop being abstract.
The Build
Built entirely on the browser's native WebCrypto API — no external crypto libraries, no backend. Covers symmetric and asymmetric encryption, hashing, HMAC, digital signatures, simulated ECDH key exchange between two parties, PBKDF2/Argon2id key derivation, JWT tooling, and in-memory file hashing/encryption. Keys and plaintext never leave the tab.
Security Decisions
AES-GCM is the default symmetric cipher because it bundles authentication into encryption itself — tampered ciphertext is caught automatically. Every operation generates a fresh IV, since reusing one with the same key silently breaks AES-GCM's guarantees. RSA-OAEP is capped to small payloads by design, which forces the correct pattern: RSA wraps a symmetric key, AES-GCM handles the actual data.
What It Exposes
Building this as a real, usable tool — not a demo — makes the constraints impossible to ignore. IV reuse, RSA's payload limits, why encoding isn't encryption: these stop being trivia and become things the tool has to actively guard against or explain.
Stack
WebCrypto API · Vanilla JavaScript · Client-side only, no backend
Read the case study
What I Set Out To Learn
The practical layer underneath cloud security — networking, Linux, and offensive-security tooling — not just theory. Completed as part of a structured Cyber Security course (CompTIA Security+ track, via 101labs.net).
Labs & Experiments
30 labs across three areas: network diagnostics (traceroute, ping, nslookup, dig, ifconfig/ipconfig, netstat, route, ARP), offensive/recon tooling (Hping, Scanless, dnsenum, Hashcat password cracking, MAC spoofing, WPS attacks via Reaver and Wifite, HTTP credential capture with Wireshark, packet capture with tcpdump), and scripting from fundamentals to advanced (Bash and Python).
Tools Used
A self-configured Kali Linux VM, Wireshark, tcpdump, Hashcat, Reaver, Wifite, and the native Linux/Windows networking toolset.
What I Learned
Seeing an attack from the tool's-eye view — capturing credentials sent over plain HTTP, cracking a weak hash with Hashcat — makes the reasoning behind security controls concrete in a way reading about them never does.
Stack
Kali Linux · Wireshark · tcpdump · Hashcat · Bash · Python
♃ Software & Product
Read the case study
The Idea
Prove I could design and build a fintech product end-to-end — not a mockup, a fully functional wallet with real interaction logic, positioned as "Wealth, Elevated."
The Build
A PWA, installable to the home screen. Multi-step onboarding, a tiered rewards system (Silver/Gold/Platinum), transfers, bill payments, a tap-to-flip virtual card, savings goals, and a quick-loan simulator with live repayment calculation — across six visual themes.
Offline-First Decisions
Every balance and transaction lives in the browser's local storage — nothing hits a server. That constraint shaped the architecture: every action had to update state instantly and persist locally instead of round-tripping to an API.
UX Decisions
PIN-gated transactions, hidden balance by default, hold-to-copy on the virtual card, streaks and cashback — small mechanics that make a wallet feel alive rather than static.
What I'd Improve
Swap local storage for a real backend — auth and a persisted ledger — to move it from a convincing simulation to a deployable product.
Stack
PWA · Service Worker · Vanilla JavaScript · LocalStorage
Read the case study
The Idea
A small, fun consumer app — proof I can ship outside the security/fintech lane and still make it feel considered rather than thrown together.
The Build
Browse pickup lines by category, save favorites, rate and share — a simple loop, built with real product polish: custom app icon, installable, fully themed.
UX Decisions
Kept the interaction loop tight — browse, react, save — no unnecessary steps. A "fun" app dies the moment it feels like effort to use.
Stack
Next.js · Responsive UI · PWA basics
Read the case study
The Client
Vybratez Entertainment — a Nigeria-based solo MC, host and hype man running events, artist promotion and sound equipment rental.
The Problem
A one-person business fielding bookings and questions ad hoc over DMs, with no single place to show services, credibility or availability.
The Solution
A full site: sharp positioning on the hero, a service catalog covering MC work, hyping, event planning, artist promotion, show organization, sound rental and shoutouts, booking cards per service, and an FAQ answering the questions clients actually ask before booking.
Booking Flow
Every booking path routes to WhatsApp pre-filled with context — the fastest path to a reply for a business that already runs on direct messaging, instead of a contact form that goes quiet.
Result
Live, deployed, and in active use — built and shipped for a real paying-client relationship, not a self-directed exercise.
Stack
React / Next.js · WhatsApp Deep-Linking · Vercel
♃ AI-Assisted Builds
Read the case study
The Idea
Résumé builders are almost universally template-generic. I wanted one that felt editorial — considered, typographic, like something a magazine would produce — not another dropdown-driven form.
What I Built
A builder and settings flow for composing a résumé with an editorial visual language instead of a generic corporate template.
Where AI Fits
Built on an AI-native app platform (Base44), which handled a lot of the scaffolding — but the product direction, the editorial positioning, and the visual and UX decisions were mine to shape and iterate on. AI as a build accelerant, not the whole idea.
Design Decisions
Typography-led layout choices over generic form-builder UI — the goal was for the résumé itself to look considered, not just the tool that made it.
Stack
Base44 (AI App Platform) · Editorial UI Design
♃ VISUAL WORK
Design plates
