About me

I am a Ph.D. candidate in Artificial Intelligence at École de technologie supérieure (ÉTS) in Montreal, Quebec, Canada, conducting research at LIVIA. My current research spans LLM agents, computer-use agents, agent evaluation, and vision-language models.

I was recently a visiting researcher at ServiceNow AI Research, where I worked on self-improving web agents that learn reusable skills, workflows, and memories from interaction trajectories without fine-tuning. This line of work asks when online memory and skill modules are actually worth their inference-token cost.

During my Ph.D., I have also worked on locality-aware pre-training for ViT-based models, open-vocabulary semantic segmentation, few-shot adaptation of vision-language models, and generalized few-shot semantic segmentation. I hold a master’s degree in artificial intelligence from Sharif University of Technology, where I worked on feature disentanglement in variational autoencoders (VAEs). I also completed my bachelor’s degree at the same institution.

In addition to my academic work, I have industry experience as a software engineer at Divar, a major classified ads platform in my home country.