Abdul Rauf — Portfolio — zsh
> Initializing runtime environment...
> Loading project manifests...
> Connecting to creative engine...
> Compiling experience modules...
> All systems operational.

Full-Stack Developer & AI Engineer

Abdul
Rauf.

CS @ FAST NUCES IslamabadIslamabad, Pakistan

I craft scalable web applications, multi-agent AI systems, and software that actually ships. Open to internships and freelance work.

View My Work Download CV Get In Touch

rauf17LinkedInEmailWhatsApp

scroll

01 — About

Who I am.

Hey! I'm Abdul Rauf, a CS student at FAST NUCES Islamabad (graduating 2027) with a CGPA of 3.01. I build things across the full stack — from polished Next.js frontends to ML pipelines to systems-level C++ projects.

I've interned as an AI Engineering Intern at AI Genmat, worked as a Teaching Assistant for Data Structures and OOP at FAST NUCES, and shipped real client work on Fiverr including a live medical payment portal.

Passionate about developer tooling and AI-powered applications. When I'm not coding, I'm probably optimizing my workflow or building something no one asked for but everyone needs.

15+

Public repositories on GitHub

3.01

CGPA at FAST NUCES

45%

ML pipeline speedup at AI Genmat

99.9%

API transaction reliability

02 — Experience

Where I've worked.

Jun – Aug 2025Remote

AI Engineering Intern

AI Genmat

→Engineered automated ML data pipelines using NumPy and Pandas, reducing preprocessing time by 45%.
→Implemented supervised learning models with Scikit-learn, improving baseline accuracy by 30% for research prototypes.

Aug 2025 – Jan 2026On-site

Teaching Assistant — OOP

FAST NUCES Islamabad

→Led live coding labs for 120+ students on C++ and Java OOP concepts.
→Reduced average student debugging time through structured mentoring sessions.

Jan – May 2025On-site

Teaching Assistant — Data Structures

FAST NUCES Islamabad

→Designed and evaluated a full-scale C++ project for 450+ students.
→Established grading pipelines and comprehensive evaluation rubrics.

03 — Projects

Things I've built.

◈

Featured

Multi-Agent AI Research Assistant

Production-grade multi-agent research system built with LangGraph and FastAPI. Features autonomous web search, data validation, and Human-in-the-Loop interrupts for complex research workflows.

LangGraphFastAPIPythonMulti-AgentHITL

⬡

Trevo — Dev Productivity Tool

Modern developer tool for visualizing and exporting clean directory structures across 25+ tech stacks. Linear-inspired UI with live search and real-time ASCII generation.

Next.js 15TypeScriptTailwindFramer Motion

◎

Readme-AI

AI-powered documentation engine transforming raw GitHub repos into professional README files using Gemini 2.5 Flash. Code Rain glassmorphic UI, live editor, syntax highlighting.

Next.jsGemini APITypeScript

◻

ApexDocs — Document SaaS

SaaS platform for converting Markdown into professional PDFs. Firebase Auth, Firestore persistence, real-time markdown editor with a minimalist Tailwind UI.

ReactViteFirebaseTailwind CSS

◯

Orbit — Timezone Visualizer

High-precision 24-hour timeline for mastering global timezones. Visualize daylight overlaps, track team availability, coordinate across the globe with a minimalist interface.

TypeScriptNext.jsProductivity

⬜

Client

AnimalTriage — Medical Portal

Integrated full Stripe payment lifecycle into a live medical portal. Debugged 250+ critical API errors, achieving 99.9% transaction reliability. Supabase webhook automation for patient workflows.

StripeSupabaseWebhooksFiverr

◆

FlexPrime — FAST Portal Extension

Chrome extension supercharging the FAST-NUCES FlexStudent portal with marks alerts, GPA simulator, attendance tracking, auto grand total fix, and fully redesigned UI. 41 GitHub stars.

JavaScriptChrome ExtensionFAST NUCES

◉

Chrono Rift OS

Multi-process C++17 Linux game built for OS concepts — POSIX shared memory, semaphores, pthreads, ncurses TUI, and signal handling across three concurrent processes.

C++17POSIXpthreadsncurses

▣

Hand Gesture Volume Controller

Real-time computer vision app controlling system volume through hand gestures via OpenCV and MediaPipe. Streamlit web interface, no local setup required.

PythonOpenCVMediaPipeStreamlit

◈

CUDA KLT Tracker

GPU acceleration of the KLT Feature Tracker using CUDA (HPC course). Sequential CPU baseline (V1) and progressively optimized GPU implementations (V2–V4).

CUDACGPUHPC

04 — Skills

What I know.

Languages

TypeScriptJavaScriptPythonC++JavaSQLCUDAx86 ASM

Frontend

React.jsNext.jsTailwind CSSFramer MotionHTML5CSS3

Backend

Node.jsExpress.jsFastAPIREST APIsWebhooks

Databases

MongoDBPostgreSQLMySQLSupabaseFirebase

AI / ML

TensorFlowPyTorchScikit-learnLangGraphOpenCVNumPyPandas

Tools

GitDockerVercelStripeJupyterStreamlit

Certifications

◎Google AI Essentials Specialization ◈AI For Everyone — DeepLearning.AI ◆Intro to Generative AI — Google ◻EF SET English Certificate — C2 Proficient ▣Linux Shell Scripting for DevOps

05 — Contact

Let's talk.

Open to internships, freelance work, and interesting collaborations. Whether you have a project in mind or just want to say hi — my inbox is always open.

Connect Across Platforms

GitHub

AbdulRauf.