Building AI that works in the real world — not just in notebooks.
I am an AI/ML Engineer with production experience delivering intelligent systems for clients across the US, Malaysia, and Turkey. My work spans computer vision, large language models, vision-language models, and edge AI deployment on constrained hardware.
I do not just run experiments. I own pipelines end to end, from data annotation and model training to production deployment. I am also an active researcher at FAST-NUCES with contributions in Explainable AI, IoT security, and NLP progressing toward publication.
Deeply interested in Quantum AI and its intersection with computer vision and healthcare, a space I believe will define the next generation of intelligent systems.
"I take a good idea, go build it, and come back with receipts."
Computer Vision → Object Detection · SAR Imagery · Satellite Analysis · Medical Imaging
LLMs & Agents → RAG · Graph RAG · Fine-tuning · Prompt Engineering · Voice Agents
Vision-Language → BLIP-2 · VLMs · Video Intelligence · Image Captioning
Edge AI → Qualcomm NPU/GPU/CPU · On-device LLM · Android Inference · Quantization
Healthcare AI → Clinical RAG · Doctor-Patient Transcription · Medical NLP
| Project | Domain | Stack | Client |
|---|---|---|---|
| Energy Asset Detection System | SAR Imagery / Energy | YOLO, CVAT, VV/VH Polarization | 🇺🇸 US Energy |
| Aerial Storage Infrastructure Classifier | Aerial Computer Vision | YOLO11, CVAT, RTX 4070 | 🇺🇸 US Energy |
| Geospatial Infrastructure Intelligence Pipeline | Geospatial AI | YOLO, MAXAR, VEXCEL, QGIS, OSM | 🏢 Enterprise |
| Enterprise Document Intelligence Platform | Fintech Document Intelligence | Neo4j, Python, FastAPI | 💰 Fintech |
| Retail Loss Prevention Vision System | Video Intelligence | YOLO, LSTM, VLMs | 🇲🇾 Malaysia |
| Multi-Domain Conversational AI System | Conversational AI | Django, LLMs, LangChain | 🇹🇷 Turkey |
| Autonomous Voice Interaction Agent | Agentic AI | VAPI, LangFuse, OpenAI | 🏢 Enterprise |
| Intelligent Document Processing Engine | OCR & Parsing | DeepSeek OCR, PaddleOCR, Azure | 🏢 Enterprise |
| Medical Consultation Intelligence Platform | Healthcare AI | Next.js, LlamaIndex, OpenAI | 🏥 Healthcare |
| Project | Tech | Highlights |
|---|---|---|
| Clinical RAG — MIMIC-IV | ChromaDB, Gemini 2.0, Streamlit | Zero hallucination · Perfect faithfulness |
| BLIP-2 Fine-tuning (LoRA) | HuggingFace, PEFT, PyTorch | BLEU 0.337 · CIDEr 0.929 |
| Android On-Device YOLO | Java, YOLO26n, Qualcomm Tools | Real-time inference · Zero cloud dependency |
| Multi-Camera Detection | Python, OpenCV, YOLO26n | 3 simultaneous streams · No CUDA needed |
| Adaptive ACL | Solidity, SDN, Game Theory | IoT security · Blockchain ACL · Targeting publication |
| LLM XAI Benchmarking | Transformers, XAI Frameworks | 5+ SOTA LLMs compared · Targeting publication |
| Urdu Sentiment Analysis | BERT, TF-IDF, Word2Vec | 90% accuracy · Low-resource NLP |
| Transformer: Pseudocode → C++ | PyTorch, Kaggle GPU | Seq2seq code generation |
| DeepSeek Math Fine-tuning | PEFT, LoRA, HuggingFace | Math reasoning · Riddle solving |
AI/ML
LLMs & Agents
Edge AI
Backend & Tools
Data & Visualization
- 🥉 Cum Laude
- 📋 Dean's List — 8 Consecutive Academic Terms
- 🏆 ICPC Qualifier — 2021 & 2025
- 🎓 Leading Edge Skills American Scholarship — Data Analytics ($2,750)
I am open to remote opportunities, research collaborations, and ambitious AI projects worldwide.
If you are building something hard and meaningful in AI — I would love to connect.