~/rafay-adeel

$ whoami

Rafay Adeel

Full stack & automation engineer working across web apps, AI/ML and computer vision, and workflow automation. BS Computer Science from FAST NUCES, cum laude, CGPA 3.68.

now
Full Stack & Automation Engineer at TheBrandSpur, Faisalabad
open to
full-time roles · onsite in Faisalabad, Lahore, Islamabad or abroad · remote
email
rafayadeel1999@gmail.com
github
@CodeRafay
linkedin
in/rafay-adeel

press / for commands

README

Rafay Adeel is a software engineer from Faisalabad, Pakistan. He graduated cum laude from FAST NUCES in June 2026 with a BS in Computer Science (CGPA 3.68), and works as a Full Stack & Automation Engineer at TheBrandSpur, where he builds APIs, integrations, dashboards and AI-powered automation for supply-chain operations.

His work spans full stack development (React, Next.js, Node.js, Flask, FastAPI), AI/ML and computer vision (PyTorch, OpenCV, YOLOv8, RAG pipelines) and infrastructure (SLURM, Docker, Prometheus). Before that he interned at Nexium Global and designed AI training data for Turing. He is open to full-time roles, onsite in Faisalabad, Lahore, Islamabad or abroad, or remote.

Longer term, he wants to move toward research: he's interested in computer vision and applied machine learning, and hopes to pursue graduate studies.

SELECTED WORK

[01]

HPC-Slurm-Hex infra

A browser GUI for SLURM clusters. It provisions nodes with Ansible, handles job submission and user management with RBAC, scans submitted code for malicious patterns, and uses Prometheus and Grafana for monitoring. Researchers can run jobs without SSH.

stack React · Express · Docker · Ansible · Prometheus · Grafana

[02]

Veritas Forensics vision

A forensic inspector that runs 13 computer-vision techniques on a suspect image, including Error Level Analysis, PRNU sensor-noise fingerprinting, FFT/DCT frequency analysis, copy-move and resampling detection, and deepfake artifact checks. It shows the evidence and leaves the verdict to the analyst. It's my most-starred repo.

stack Python · OpenCV · Streamlit

[03]

Self-Correcting RAG llm

Retrieves context from PDFs, has an LLM grade each chunk for relevance, and falls back to web search when retrieval is weak. This keeps answers grounded in sources instead of hallucinated.

stack FastAPI · React · Pinecone (MMR) · HuggingFace embeddings

[04]

img2md vlm

Turns photos of documents into clean Markdown. It fine-tunes Qwen2-VL with QLoRA adapters, pulls the base model from Hugging Face at runtime, and ships only the small adapter weights.

stack Qwen2-VL · QLoRA · PyTorch · Streamlit

[05]

Ephemeral P2P Chat web

One-to-one chat where messages go browser to browser over WebRTC DataChannels. A Cloudflare Worker handles signaling only. There is no database, no relay and no message history.

stack WebRTC · Cloudflare Workers · Next.js · TypeScript

[06]

Urdu2Roman nlp

Transliterates Urdu script into Roman Urdu. It uses a BPE tokenizer written from scratch and a BiLSTM encoder with an LSTM decoder, and is evaluated with BLEU and CER.

stack PyTorch · custom BPE · Streamlit

$ sbatch --demo

A toy version of what HPC-Slurm-Hex manages. Submit jobs and watch a FIFO scheduler pack them onto four 4-core nodes.

queue 0 · running 0 · done 0

    MORE WORK

    RECORD

    experience

    aug 2026 — now
    Full Stack & Automation Engineer, TheBrandSpur · onsite, Faisalabad
    • Automates data imports from APIs, FTP feeds and databases, and builds inventory models (reorder points, safety stock, EOQ, ABC analysis) and demand forecasts.
    • Designs REST and GraphQL APIs with OAuth 2.0 / JWT, and connects e-commerce platforms, ERPs, 3PLs and warehouse systems through webhooks with retries and error handling.
    • Builds KPI dashboards, supplier scorecards and threshold alerts (low stock, shipment delays), with reports sent automatically to email, Slack and Teams.
    • Integrates Claude and OpenAI APIs into ops chatbots and NLP workflows for email classification, contract clause extraction and document summarization, and supports forecasting and recommendation models.
    sep — nov 2025
    AI Conversation Designer (freelance), Turing · remote
    • Designed multi-turn datasets for training AI assistants in reasoning, tool use and realistic dialogue.
    • Wrote system prompts and conditional multi-step tool workflows, and ran failure analysis and adversarial tests to harden models.
    jul — aug 2025
    AI-First Web Development Intern, Nexium Global Ltd · remote
    • Built REST APIs in Node.js and Flask for AI-powered features, and backend modules for data processing.
    • Integrated responsive frontend components with backend services.

    education

    2022 — 2026
    BS Computer Science, FAST NUCES Faisalabad · graduated June 2026
    Cum laude · CGPA 3.68 · Dean's List Fall '22 to Fall '25 · Silver Medal (Fall '22)
    teaching
    Teaching Assistant, FAST NUCES
    Programming Fundamentals · OOP · Linear Algebra · Software Design · Database Systems

    certifications

    ibm
    Python for AI & Data Science
    google
    Cybersecurity · Project Management · Data Analytics · IT Support

    leadership

    2023 — 2025
    TEDxFAST Faisalabad · Event Lead, Daira '25 · Management, Daira & Job Fair '23–'25

    TOOLBOX

    languages

    • Python
    • JavaScript
    • TypeScript
    • C++
    • C
    • C#
    • SQL
    • Bash

    full stack

    • React
    • Next.js
    • Node.js
    • Express
    • Flask
    • FastAPI
    • GraphQL
    • Tailwind CSS

    ai / ml / cv

    • PyTorch
    • TensorFlow
    • OpenCV
    • scikit-learn
    • Hugging Face
    • ONNX Runtime
    • YOLOv8
    • Pandas
    • NumPy
    • Streamlit

    automation

    • Apps Script
    • Google Sheets
    • OpenAI API
    • Claude API

    data

    • PostgreSQL
    • MongoDB
    • SQLite
    • Oracle
    • Supabase
    • Prisma

    infra

    • Docker
    • Kubernetes
    • AWS
    • Git
    • GitHub
    • Ansible
    • Prometheus
    • Grafana

    FAQ

    Are you available for work?

    Yes. I'm open to full-time roles in full stack, AI/ML, computer vision and automation. I can work onsite in Faisalabad, Lahore, Islamabad or abroad, or remotely.

    What do you work on?

    Full stack apps and APIs, automation and integrations for business operations, and AI/ML: computer vision (forensics, detection, reconstruction) and LLM systems that stay grounded in sources (RAG, vision-language models).

    What's your stack?

    React/Next.js, Node, Flask and FastAPI for the web; PyTorch, OpenCV and Hugging Face for ML; Apps Script, webhooks and LLM APIs for automation; Docker, Kubernetes, AWS and Prometheus/Grafana for infrastructure. The full list is in the toolbox.

    Are you interested in research?

    Yes. Longer term I want to work on research in computer vision and applied machine learning, and I hope to pursue graduate studies.

    How do I reach you?

    Email rafayadeel1999@gmail.com or use the form below. I usually reply within a day or two.

    CONTACT

    rafayadeel1999@gmail.com

    Faisalabad, Pakistan (PKT, UTC+5)