Hossameldin Mohammed Email & Phone Number

Building AI model-centric microservices with in-house built models, and LLM RAG pipelines. End-to-end full stack development of machine learning solutions for content generation, NLP tasks like semantic behavioral tagging, and customer segmentation.

Burnaby, British Columbia, Canada

Building and deploying ML systems on edge devices for applications like:- Touch-screen stylus latency recognition system: using a GRU-CNN architecture, incorporating temporal sequence analysis.- Online handwriting recognition: using seq-to-seq autoencoders, for over 60 languages with 96% accuracy, leveraging NLP methods for character and language modeling.- Image matting system: using a modified YOLO and U2NET architectures, with a focus on semantic segmentation.- Off-line signature fraud identification system: using dynamic signature synthesis and verification (synsig2vic).- Touch-screen virtual smart keyboard: integrating NLP predictive text and gesture recognition.

Vancouver, British Columbia, Canada

Berkeley, California, United States

Building NLP-based Email Recommendation Engine that includes image, sentiment analysis, word count, send-time and quantitative prediction and optimization.

Burnaby, British Columbia, Canada

Teaching machine learning and software engineering courses, in addition to leading small student research projects.

Vancouver, Canada

Conducting research on Imitation Learning generative modeling for traffic systems. Building LSTM-based Variational Autoencoders, clustering of cyclist movements, and collaboration in the development of BITSAFS Traffic Intelligence (TI) software for computer vision-based traffic analysis.

Cairo Governorate, Egypt

Working on data-driven approach for sustainable mobility, traffic management programs, building traffic data warehousing systems, truck sampling strategies for a portable emission measurement study.

Conducting research on modeling and simulation of large-scale transportation network systems.