Muhammad Habeel
Building intelligent systems that See · Understand · Automate.
Specialized in multi-modal biometric verification, anti-spoofing liveness pipelines, sub-second vector similarity search, and high-performance FastAPI backends for real-world AI deployment.
What I Build
End-to-end engineering from deep neural network training to scalable cloud deployment.
Computer Vision
Multi-modal facial recognition, secondary iris pattern matching, EfficientNet anti-spoof liveness models, and deepfake artifact detection.
NLP & Conversational AI
Domain-specific medical chatbots, clinical symptom triage, Hugging Face transformers, and semantic embeddings for dialogue systems.
AI Automation & Pipelines
High-throughput vector indexing (FAISS), asynchronous FastAPI pipelines, edge optimization with ONNX, and automated inference services.
Full-Stack Systems
Hardened Node/Express and Next.js applications, robust security middleware, clean API contracts, and high-performance user experiences.
Featured Projects
BioAttend
Multi-modal biometric attendance — face, iris, and liveness in one pipeline
Built a FastAPI-based multi-modal verification pipeline combining InsightFace Buffalo_L embeddings + FAISS for fast face search, iris verification as a second factor, and a custom EfficientNet-B0 liveness model trained on a self-collected video dataset to reject spoof attempts.
Lahore TravelMate
A Node/Express tourism platform for exploring Lahore
Built and am currently hardening a Node/Express tourism site — closing security gaps and polishing UX for public release.
Medical AI Research
University research across medical imaging, health chatbots, and wildfire detection
Conducted university research projects covering medical imaging analysis, a nasal health chatbot, and a wildfire detection model.
DeepFake Detection
Detecting manipulated faces using transfer learning
Built a classification notebook using Kaggle's 140k real-vs-fake faces dataset, applying MobileNetV2 and ResNet50 transfer learning to detect manipulated images.
Experience
Key Responsibilities & Deliverables
Architecting production-ready AI pipelines spanning biometrics, face embeddings search, and real-time anti-spoofing liveness verification.
Developing low-latency inference services with FastAPI, ONNX Runtime, and TensorRT for edge and cloud deployment.
Building intelligent automation agents and multimodal retrieval pipelines integrating LLMs, vector databases (FAISS, Chroma), and custom vision backbones.
Collaborating with cross-functional engineering teams to harden production systems, optimize GPU compute, and maintain robust API contracts.
Let's build something intelligent.
Have a computer vision problem, biometric workflow, or deep learning system you want to bring to life? Let's connect and architect solutions together.