Smoke & Fire Detection
Real-time detection pipeline with TensorRT optimization and inference target of 100ms or less.
I build production-focused AI systems for computer vision, edge deployment, real-time inference, automation, and modern LLM workflows.
I specialize in computer vision and deep learning with a strong focus on speed, optimization, and deployment. My work spans object detection, real-time inference, TensorRT acceleration, RAG pipelines, LLM fine-tuning, APIs, and automation.
Computer vision systems with YOLO, OpenCV, TensorFlow, and PyTorch
Edge deployment on Jetson Nano and Raspberry Pi
Model optimization with TensorRT, FP16, and INT8 quantization
NUST CEME, Islamabad. Coursework in deep learning, computer vision, NLP, embedded AI, cybersecurity, AI, and ML.
Real-time inference, GPU acceleration, optimized deployment pipelines, and practical embedded AI builds.
Smoke and fire detection, Whisper speech-to-text, breast cancer detection, and smart parking detection.
Computer vision, object detection, CNNs, model optimization, and real-time AI workflows.
My skill set combines AI research fluency with developer habits: clean Python, measurable evaluation, API thinking, deployment awareness, and optimization for real-world constraints.
Production-focused systems
YOLO, OpenCV, CNNs
Jetson Nano, Raspberry Pi
RAG and automation
NUST CEME, Islamabad
Graduate coursework
Graduate coursework
Graduate coursework
Real-time detection pipeline with TensorRT optimization and inference target of 100ms or less.
Real-time transcription workflow supporting microphone input and uploaded audio files.
Classification model with structured evaluation for healthcare-oriented ML experimentation.
Computer vision system for detecting parking occupancy and supporting real-time space monitoring.
Retrieval-augmented workflows for scalable context search, response generation, and automation.
Acceleration experiments using GPU workflows, FP16/INT8 quantization, and pipeline tuning.
Tools I use to move AI projects from experiments to working software.
Send a short note with your goal, data type, timeline, and deployment target. I can help shape the model, pipeline, evaluation, and production path.
Email Maham GitHub