Machine Learning Engineer | Computer Vision & ML Systems
I build practical machine learning systems with a focus on computer vision, deep learning, inference pipelines, APIs, and deployment.
My approach is to take ML projects beyond experimentation: from model integration and evaluation to post-processing, API serving, optimization, and deployment.
BiRefNet-based background removal with custom structural foreground analysis, adaptive alpha refinement, and fine-detail processing.
Highlights
- Pretrained BiRefNet foreground segmentation
- Soft-mask preservation
- Image-aware mask refinement
- Structural foreground analysis
- Adaptive alpha refinement
- Hair and fine-detail processing
- FastAPI inference API
- Gradio interface
- Hugging Face Spaces / ZeroGPU deployment
Live Demo: Hugging Face Space
Machine Learning & Deep Learning
Python · PyTorch · Torchvision · NumPy · Pandas
Computer Vision
OpenCV · Image Segmentation · Object Detection · Landmark Detection · Image Processing
ML Engineering & Deployment
FastAPI · Gradio · Hugging Face Spaces · ZeroGPU · ONNX · REST APIs · Model Inference
Speech, NLP & Automation
Whisper · spaCy · pyttsx3 · SQLite · APScheduler · Playwright
Development Tools
Git · GitHub · Docker · VS Code · Google Cloud
- Computer Vision
- Deep Learning
- ML Inference & Optimization
- Model Deployment
- Multimodal AI
- Generative AI
- MLOps
- AI Agents
- Building production-oriented ML systems
- Improving model inference and deployment workflows
- Developing a 3D eyewear virtual try-on system
- BS Computer Science — The Islamia University of Bahawalpur (2025)
- Practical experience across computer vision, speech/NLP, automation, and ML deployment
- Professional training in Machine Learning & Deep Learning
- Applied ML engineering experience through internship work on computer vision systems
- GitHub: github.com/RazaAli9
- LinkedIn: linkedin.com/in/ali-yousaf783
- Hugging Face: huggingface.co/RazaAli89
From model to production — building practical machine learning systems end to end.