Mehrdad Zakershahrak

Projects

Project 1

DEIO Robotics Hardware Engineering

A kid-size humanoid robot I engineered for RoboCup competitions, ranking 3rd in the U.S., featuring dynamic locomotion, real-time vision tracking, and agile autonomous soccer play.

Project 1

PUF Hardware Engineering

A two-way multimedia authentication system built on FPGA using physically unclonable functions (PUFs), enabling secure identity verification through unique, tamper-proof hardware signatures.

Project 1

OEG AI Robotics

I developed online explanation generation (OEG) for human-robot teaming, reducing cognitive load by incrementally delivering model-reconciling justifications during task execution. Evaluated across domains with strong empirical results

Project 2

PEG AI Robotics

I introduced progressive explanation generation (PEG) for human-robot teaming, modeling explanation order preferences via inverse reinforcement learning to reduce cognitive load and improve interpretability during task execution.

Project 3

NOVA Robotics AI

I built a VR-guided robotic system that simulates and validates manipulation plans before execution. By previewing actions in virtual space, it ensures safety, reduces risk in high-precision tasks, and enables intuitive human-in-the-loop control.

Project 4

MOSAIC AI DevOps

I developed a web-based AI facilitation platform for distributed teams managing biased, asymmetric information. It guides conversations to reduce noise, surface shared context, and support aligned, robust decision-making.

Project 4

Fixo DevOps AI

I launched an agentic AI tech support service on MS Azure that runs 40+ traditional tools to resolve issues end-to-end. Minimalistic, fast, and autonomous—built to deliver seamless, 24/7 technical assistance at scale.

Project 4

Linkedfull DevOps AI

I designed and built a next-generation fashion commerce platform that reimagines how users discover and engage with elegant fashion. Moving beyond traditional e-commerce, I developed an interactive, algorithm-driven shopping experience that replaces endless scrolling with intelligent style discovery—empowering users to explore with confidence, clarity, and joy.