Name: Musab Ahmed Khan Umair
Location: Istanbul, Turkey
Education:
- MSc Artificial Intelligence — Istinye University (2025 – Present)
- BE Computer Engineering — Istinye University (GPA: 3.17, Graduated Feb 2025)
Current Role: AI Research Intern @ TUBiTAK & Yuva Intelli AI Solutions
Focus: AI/ML Research | Medical Imaging | NLP | Full-Stack Development- 🎓 MSc in Artificial Intelligence at Istinye University, Istanbul
- ❤️ TUBiTAK Research Intern — Pediatric Cardiology Anomaly Detection (U-Net, MRI segmentation)
- 🤖 AI Research Intern — OCR systems, Arabic audio analytics (Whisper, CAMeL-BERT)
- 🩺 Published research: 99.53% accuracy in fetal health analysis using Random Forest
- 🚀 Full-stack builder — .NET, React, Kotlin, Supabase, Docker
- 🌏 Based in Istanbul, Turkey | From Hyderabad, India
|
Research Intern — Pediatric Cardiology Anomaly Detection Istinye University, Istanbul | Dec 2024 – Present 36-month project using U-Net segmentation on MRI scans (100+ patients) to detect congenital heart anomalies via geometric analysis and computational modeling. |
|
|
AI Research Intern Yuva Intelli AI Solutions, Hyderabad | Feb 2025 – Present Built OCR system (OpenCV + Tesseract + Azure + Gemini). Developing Arabic audio analytics pipeline (Whisper + CAMeL-BERT + LDA/TF-IDF). |
|
|
Student Research Assistant Istinye University | Nov 2023 – Feb 2025 ML in healthcare — hyperthyroid detection (1,356 patients), fetal health analysis (2,126 patients, 99.53% accuracy with Random Forest). |
|
|
Computer Engineer Intern Big Property Agency, Istanbul | Jul – Aug 2023 Built real estate website (30+ properties, JWT auth, MERN stack) with 5 investment guides. |
|
Automatic image segmentation model for early-stage pediatric cardiology anomaly detection using geometric analysis on MRI scans. |
Arabic audio pipeline — transcription, sentiment analysis, and topic modeling with interactive visual dashboards. |
|
Full-stack Kotlin Android app (MVVM, Hilt, Compose) with AI-powered Turkish speaking coach, real-time speech-to-text feedback, and offline caching. |
Full-stack recipe tracker for 100+ home-cooked meals with tag-based filters, JWT auth, image uploads via Cloudinary, deployed on Render. |
|
ML research on 2,126 patient records achieving 99.53% accuracy for fetal health risk classification using optimized Random Forest. |
End-to-end OCR system with multi-engine text extraction (Tesseract, Azure, Gemini) and automated batch number detection with confidence logging. |
⚡ AI Researcher by day, Full-Stack Builder by night ⚡
"From MRI scans to mobile apps — engineering solutions that matter."