About Hi, I'm Asiye
đź“„ Download My Resume
Who I am and What I do?
I am an experienced Machine Learning Scientist and AI Engineer with a strong background in deep learning, graph neural networks, and predictive analytics. Passionate about AI-driven problem-solving, I specialize in developing cutting-edge models for NLP, generative AI, and intelligent systems. My work spans from academic research to industrial applications, where I deploy scalable AI solutions on cloud platforms. With a PhD in Information and Systems Engineering and multiple publications in top AI conferences, I thrive on pushing the boundaries of ML research while mentoring aspiring data scientists.
Let’s innovate together!
How I Got Here
My journey in AI and data science began with a strong foundation in transportation analytics, where I applied data-driven solutions to optimize mobility and urban planning. As a Data Scientist | Transportation Planner at Iriana Co, I worked on predictive modeling and optimization, leveraging machine learning to solve real-world challenges.
Seeking to deepen my expertise, I transitioned to the Traffic Lab at IUST, where I worked as both a Data Scientist and Researcher. While my focus expanded into research, I continued developing machine learning models for transportation systems, improving forecasting, demand analysis, and network optimization. This phase strengthened my ability to bridge practical applications and research-driven AI innovations.
In 2021, I took a major step forward by joining BusPas Inc. in Montreal as a Machine Learning Scientist | AI Engineer, focusing on deep learning, AI-powered mobility solutions, and scalable machine learning models. This role allowed me to integrate cutting-edge AI techniques into real-world transportation problems, pushing the boundaries of intelligent mobility solutions.
Currently, I am engaged in advanced AI research and development, working with a wide range of models, including Graph Neural Networks (GNNs), attention mechanisms, transformers, computer vision models, LSTMs, reinforcement learning, Generative AI, and federated learning. My work is not limited to transportation—these models have applications across multiple domains, from healthcare and finance to autonomous systems and beyond.
Beyond research and development, I am also passionate about mentoring and advising students, researchers, and aspiring AI professionals, fostering the next generation of AI talent. I thrive on exploring innovative ways to apply AI to solve complex problems and drive meaningful impact across industries.
My Expertise & Contributions
Graph Neural Networks (GNNs)
GNNs enable AI to process graph-structured data, making them essential for applications like social network analysis, recommendation systems, and traffic prediction. I specialize in developing attention-based GNNs, GNN-LSTMs, and scalable architectures for complex systems, ensuring efficient modeling of spatial and temporal dependencies.
Generative AI & Transformers
I work with Generative AI, Transformers, and Multimodal AI to enhance machine understanding and decision-making. From LLMs for intelligent query processing to transformers for sequence modeling, I integrate advanced deep learning techniques across domains. With expertise in attention mechanisms and representation learning, I continuously explore innovative AI solutions.
Spatial-Temporal Modeling
Dynamic systems require models that capture both spatial and temporal dependencies. My expertise in CNNs, LSTMs, and hybrid deep learning architectures allows me to build AI models for time-series forecasting, anomaly detection, and spatiotemporal event prediction in transportation, healthcare, and finance.
Reinforcement Learning (RL)
RL empowers AI to learn through interaction and feedback. My work in policy optimization, deep Q-learning, and multi-agent RL applies to robotics, autonomous systems, and decision-making frameworks, enabling AI models to adapt and optimize in real-world environments.
Transfer Learning & Pretraining
AI models are more effective when they learn from existing knowledge. I apply transfer learning and pretraining techniques to improve model generalization across tasks, leveraging architectures like domain-adaptive learning and model fine-tuning to reduce data dependency and training time.
Federated Learning & Privacy-Preserving AI
Data privacy is crucial in modern AI. My work in federated learning, differential privacy, and secure AI ensures model training across distributed datasets without compromising user privacy, making AI solutions applicable in sensitive domains like healthcare and finance.
Cloud AI & Scalable Deployment (Azure)
AI solutions must be scalable and efficient. I have experience deploying cloud-based AI solutions on Azure, optimizing machine learning pipelines, real-time inference systems, and distributed model training for seamless integration into enterprise applications.
Mentorship & Collaboration
AI thrives through knowledge sharing and teamwork. I actively mentor students, researchers, and professionals, guiding them in AI model development, research methodologies, and technical problem-solving. Fostering an innovative and inclusive AI community is a core part of my work.