AI Engineer | Specializing in Multimodal Healthcare Systems & Large-Scale Data Pipelines Transforming raw clinical data into actionable diagnostic intelligence.
I architect end-to-end ML systems that bridge the gap between raw data and clinical utility. My expertise lies at the intersection of Medical Imaging, Multimodal Learning, and Scalable Data Engineering.
- 🧬 Medical AI & Diagnostics: Specialized in multimodal fusion (Late-Fusion architectures), Vision Transformers (ViT), and Explainable AI (XAI) using Grad-CAM to ensure clinical transparency.
- 📊 Data Infrastructure: Experienced in building high-throughput scraping pipelines (500k+ records), time-series processing for wearable sensors, and ETL optimization.
- 🚀 AI Productization: Deploying production-ready models using FastAPI, Docker, and Hugging Face Spaces, ensuring seamless integration between ML backends and user-facing interfaces.
| Category | Tools & Technologies |
|---|---|
| Deep Learning | PyTorch TensorFlow Hugging Face Keras Vision Transformers |
| Machine Learning | Scikit-Learn XGBoost Pandas NumPy Imbalanced-Learn |
| Medical AI | MedCLIP Grad-CAM DICOM Multimodal Fusion TabNet |
| Data & Backend | FastAPI SQL BeautifulSoup Selenium Docker MLflow |
| Visualization | Plotly Seaborn Matplotlib Grad-CAM |
| DevOps/Tools | Git Linux VS Code Anaconda GitHub Actions |
Architecture: Late-fusion ensemble of MedCLIP (ResNet50) for imaging and TabNet for clinical data. Impact: Achieved 0.91 AUC, providing a professional hosted platform for real-time clinical inferences via FastAPI.
Architecture: Time-Series Transformers applied to wearable sensor data. Focus: Solving temporal dependencies in sensor data for high-accuracy human activity classification.
Architecture: Benchmarking Vision Transformers (ViT) vs DINOv2. Innovation: Integrated Grad-CAM to visualize model decision-making, critical for medical validation.
Focus: A comprehensive Social Engineering Awareness Training platform, demonstrating full-stack development and security education.
Architecture: Full-stack NLP pipeline with FastAPI, PostgreSQL, Celery, and AI-powered article analysis. Focus: Real-time news aggregation, sentiment analysis, and intelligence reporting.
I build high-performance, aesthetically minimal web interfaces for corporate and personal brands.
- Kontemporary Consulting Ltd: A high-authority corporate redesign focusing on "Quiet Authority" aesthetics.
- Client Portfolio: Specialized in developing hosted platforms for professional services and medical diagnostic portals.
- B.Sc. Computer Science | Landmark University
- Data Engineering Intern | Kontemporary Consulting Ltd
- Engineered a nationwide scraping pipeline covering 36 states and 500k+ records, reducing manual data collection effort by 85%.
- 📖 LLM Fine-tuning: Optimizing domain-specific models for medical literature.
- 🔗 Multimodal RAG: Combining vision and text for advanced clinical decision support.
- ⚡ Rust for ML: Exploring high-performance data processing.
Open to Junior ML Engineer / Data Scientist roles. Let's build the future of intelligent healthcare.