Alireza Rezazadeh

[ Email | Google Scholar | LinkedIn | GitHub ]

I am a Research Scientist Manager at Accenture’s Center for Advanced AI, where I develop solutions for multi-agent collaboration. My work focuses on optimizing agentic workflows and enhancing memory representations for LLMs.

I hold a PhD in Electrical and Computer Engineering from the University of Minnesota, where I focused on graph-structured representation learning for embodied AI.

Selected Publications

From Isolated Conversations to Hierarchical Schemas: Dynamic Tree Memory Representation for LLMs

Alireza Rezazadeh, Zichao Li, Wei Wei, Yujia Bao

ICLR 2025, also presented at NeurIPS 2024 (Sys2-Reasoning Workshop)

[Paper]

SlotGNN: Unsupervised Discovery of Multi-Object Representations and Visual Dynamics

Alireza Rezazadeh, Athreyi Badithela, Karthik Desingh*, Changhyun Choi*

ICRA 2024

[Project] [Paper] [Video]

Hierarchical Graph Neural Networks for Proprioceptive 6D Pose Estimation of In-hand Objects

Alireza Rezazadeh, Snehal Dikhale, Soshi Iba, Nawid Jamali

ICRA 2023

[Project] [Paper] [Video]

KINet: Unsupervised Forward Models for Robotic Pushing Manipulation

Alireza Rezazadeh, Changhyun Choi

RA-L 2023, ICRA 2024

[Project] [Paper] [Video]

Force Field Generalization and the Internal Representation of Motor Learning

Alireza Rezazadeh, Max Berniker

PLoS 2019

[Paper] [Data]

Education

Ph.D. in Electrical and Computer Engineering, University of Minnesota
Thesis: "Learning Graph-Structured Representations for Robotic Manipulation"
M.Sc. in Electrical and Computer Engineering, University of Minnesota
Minor: Economics
M.Sc. in Mechanical Engineering, University of Illinois Chicago
Thesis: "Force Field Generalization and the Internal Representation of Motor Learning"
B.Sc. in Mechanical Engineering, Sharif University of Technology