At CMU-Q's robotics lab, I'm building the autonomous navigation stack for an indoor patrol robot from the ground up: path planning, real-time localization, and recovery behaviors that let it keep working even when hardware is limited. Because the robot depends on an unreliable wireless network, I designed a distributed system architecture that self-heals, so when connections drop it reconnects automatically instead of leaving the robot stranded mid-patrol. I've also built out environment mapping and live telemetry with remote visualization, so we can see exactly what the robot sees and how it's behaving from anywhere, giving the team full situational awareness without needing to be in the room.
yousefh@cmuq:~$ whoami
Yousef Hussein
Computer Science @ Carnegie Mellon University. Building software, AI, and systems to solve real-world problems.
about
whoami
I'm Yousef, now studying Computer Science at Carnegie Mellon University. I like working close to the metal: software, AI, and systems built to solve real-world problems.
Right now I'm a Research Assistant at CMU-Q working on PatrolBot, an autonomous indoor robot, where I focus on the perception, localization, and navigation that lets it operate independently in dynamic indoor spaces.
- B.S. Computer Science @ CMU-Q
- Concentration: Computer Systems
- Minor: Mathematical Sciences
- Languages: English, Arabic, French
skills
toolbox
Languages
Frameworks & Tools
Data & ML
Platforms
experience
where I've worked
As part of OpenStack's Summer of Code, I'm contributing to Keystone-NG, a ground-up Rust reimplementation of OpenStack's identity service, where I'm rethinking how authentication and authorization work for cloud infrastructure at scale. A big part of that has been federated identity: adding OIDC integration and JWT-based login, and designing fine-grained access control that replaces the legacy role-explosion patterns the old system relied on. Since it's an open-source project with contributors spread across time zones, most of the work happens through code review, mailing-list discussions, and pull requests, and the changes I've shipped go straight into the production OpenStack Identity API (v3/v4).
At Amana Healthcare I built VR training simulations in Unity and C#, using spatial computing and physics-based interactions to recreate clinical scenarios that previously required physical mannequins, cutting reliance on them by roughly 70%. Because the modules had to run smoothly on standalone headsets, I spent a lot of time optimizing rendering and memory usage, which also brought annual hardware and maintenance costs down by about 50,000 QAR. Once the modules were ready, I led the rollout myself, running technical orientations for more than 50 staff so the hospital could adopt the new training tools without disruption.
projects
things I've built
Full-stack AI Contract Lifecycle Management platform: contract upload, automated risk analysis, semantic search, and multilingual translation. Built for the AIX Hackathon.
- Next.js
- FastAPI
- OpenAI
- Pinecone
- MongoDB
- Docker
A simulated desktop operating system built with CMU Graphics — login system, window manager, terminal, and a virtual file system.
- Python
- CMU Graphics
Real-time Conway's Game of Life simulator rendered with C++ and OpenGL.
- C++
- OpenGL
An automated ML pipeline that predicts 24-hour Polymarket asset movements from financial-news sentiment.
- Python
- Jupyter
- scikit-learn
- NLP
// more on github
- StreetFighter — A Street Fighter–style game built in Python.
- PassTheGame — Collaborative game-evolution experiment: start simple, pass it on, watch it mutate.
- github.com/ymh1874 — everything else
contact
get in touch
Open to internships, research, and interesting problems. The fastest way to reach me is email.
- email yousefh [at] andrew.cmu.edu
- email ymh1874 [at] gmail.com
- github github.com/ymh1874
- resume resume