Available for opportunities

Abdul
Mateen

AI Engineer building LLM pipelines, computer vision systems, and intelligent automation. BS Artificial Intelligence @ FAST-NUCES.

8+
Projects Built
3.23
CGPA / 4.0
5
Certifications
01 — About

At a Glance

Current Status
Year 3
BS Artificial Intelligence · FAST-NUCES Islamabad
Focus Areas
Large Language Models & RAG Pipelines
Computer Vision & Multimodal AI
Inference Optimization (CUDA)
Summary

AI student with hands-on experience building real-world systems — from LLM-powered chatbots and RAG pipelines to CUDA-optimized inference engines and computer vision tools. Comfortable working across the full stack of AI development: data preprocessing, model integration, API design, and deployment. I care about systems that actually get used, not just models that score well on benchmarks.

02 — Projects

What I've Built

02 / Research
Multi-Scale Region-Phrase Alignment (Remote Sensing VLMs)

Extended DGTRS-CLIP with a Region-Phrase Alignment module using dual-encoder contrastive learning. Achieved +3.20% relative mR and +9.01% I2T R@1 on the RSITMD benchmark. Trained on RET-2 dataset.

CLIP Remote Sensing Contrastive Learning PyTorch
03 / LLM
SmartStay AI — Hotel Front Desk Chatbot

FastAPI-based AI assistant using Qwen LLM with RAG and CRM features. Integrated a full TTS/STT pipeline (Moonshine ASR + Kokoro) for real-time voice interaction, plus Calendar, Calculator, and OpenWeather API tools.

FastAPI RAG Qwen TTS/STT
04 / NLP
TranslateSub AI — Real-Time Subtitle Translator

Chrome extension for real-time subtitle generation and live multilingual translation of streaming content. Low-latency NLP pipeline using Faster-Whisper, FastAPI, WebSockets, and VAD with sliding-window audio segmentation.

Faster-Whisper WebSockets FastAPI Chrome Extension
05 / CV + Accessibility
VoiceLens — Visual Reading Assistant

Mobile app combining text detection, OCR, and TTS to read menus, signs, and documents aloud. Designed for accessibility, helping users interpret printed text in real-world environments.

OCR Computer Vision TTS Mobile
06 / NLP
Sentiment Analysis — Restaurant Reviews

Scraped OpenTable restaurant reviews and classified sentiment (positive/negative) using LLM APIs. Added competitor analysis across restaurants with a Streamlit UI.

Streamlit LLM APIs NLP
07 / Game AI
ML Bot — Street Fighter Game AI

Trained an MLP on manually collected gameplay data from a Street Fighter-style game to make real-time predictive decisions during live matches.

MLP Reinforcement Learning Game AI
08 / Full-Stack
NASCON Event Management System

Full-stack web app for university-wide event registrations. React frontend, Node.js backend, MySQL database — with role-based access control and real-time updates.

React Node.js MySQL
03 — Skills

Technical Stack

AI / ML
PyTorch TensorFlow Scikit-learn RAG CLIP LLMs Embeddings Inference Pipelines
Programming
Python C++ C# JavaScript Node.js x86 Assembly SQL CUDA
Backend & APIs
FastAPI Flask REST APIs WebSockets Microservices
Data
NumPy Pandas ETL Pipelines Data Preprocessing Streamlit
Tools & DevOps
Git GitHub Docker AWS Azure GCP CI/CD
Compute
CUDA Multi-Stream Kaggle GPU (T4) OpenMP Parallel Computing
04 — Experience

Where I've Worked

Jan 2026 — May 2026
Lab Demonstrator
FAST-NUCES, Islamabad
  • Guided students through lab exercises, ensuring timely completion and understanding.
  • Answered technical queries and provided hands-on debugging support during sessions.
  • Evaluated lab submissions and delivered constructive feedback to improve quality.
2025 — Ongoing
Independent Research
Vision-Language Models · Remote Sensing · LLM Inference
  • Extended DGTRS-CLIP with novel Region-Phrase Alignment achieving benchmark improvements.
  • Achieved 1.26× inference speedup on EVEv2.0 via CUDA stream parallelism on dual T4 GPUs.
  • Explored fine-tuning techniques including LoRA, QLoRA, and model quantization.
05 — Education

Academic Background

Aug 2023 – Present
BS Artificial Intelligence
FAST-NUCES, Islamabad
CGPA 3.23 / 4.0 6 Semesters
2021 – 2023
FSc Pre-Engineering
Punjab College, Islamabad
978 / 1100
Certifications
IBM
Quantum Machine Learning, ML Model & Quantum Computing
Oct 2025
IBM
Data Analysis Using Python
Oct 2025
University of Florida
Artificial Intelligence
Aug 2025
IBM
Artificial Intelligence Fundamentals
Aug 2025

Let's Build Something

abdulmat40@gmail.com