ML engineer · AI researcher
Building intelligent systems that are efficient, interpretable, and useful.
I’m Sadra Berangi, a Computer Engineering graduate working across computer vision, reinforcement learning, robotics, and medical AI.
About
Research grounded in real systems
My work connects machine learning research with practical deployment—from interpretable visual models and deep reinforcement learning to gait-based clinical assessment.
I graduated from Shahid Beheshti University (National University of Iran) with a BSc in Computer Engineering. Alongside academic research, I build production computer-vision pipelines and deploy models on Linux-based embedded systems.
Research interests
Questions I care about
Computer Vision
Efficient and interpretable visual representations, lightweight architectures, and robust perception for real-world deployment.
Reinforcement Learning & Robotics
Data-efficient decision-making, learning-based control, and intelligent agents that can operate reliably outside simulation.
Medical AI
Transparent models for biomedical signals and movement data, with an emphasis on methods clinicians can understand and trust.
Selected work
Publications & manuscripts
Fuzzy-DoGNet: A Lightweight and Interpretable Deep Unstructured Neuro-Fuzzy System Based on Band-Pass Filters for Traffic Sign Recognition
S. Berangi, M. Parchami, A. Salimi
DOI ↗A Deep Reinforcement Learning Approach Combining Technical and Fundamental Analyses with a Large Language Model for Stock Trading
M. Veisi, S. Berangi, M. Shahbazi, A. Salimi
DOI ↗An Explainable Method Based on Sinc Filters to Diagnose Parkinson’s Disease Severity by Gait Cycle Analysis
A. Salimi, M. Veisi, S. Berangi
DOI ↗Experience
Research, engineering, and teaching
Junior ML & Vision Engineer
Hezar Makian Khazar
Developing computer-vision pipelines, curating large image datasets, and deploying real-time models on embedded Linux systems.
Research Assistant
Robotics and Intelligent Autonomous Agents Lab, SBU
Research across reinforcement learning, computer vision, biomedical AI, and explainable lightweight neural architectures.
Teaching Assistant
Shahid Beheshti University
Supported courses including deep reinforcement learning, robotics, artificial intelligence, signals and systems, and programming.
Let’s connect
Interested in working together?
I’m seeking fully funded PhD opportunities for 2027 and welcome conversations about research in vision, robot learning, and medical AI.
sadraberangi@gmail.com