AHMEDHASHASH
Computer engineering student at GJU. I train machine learning models, build backends, and wire up hardware, like a 3D-printed hand that copies your fingers.
every layer below is a tool you can use
ML & Data
I design predictive machine learning models and apply core AI principles to process and interpret data.
Then I check whether they're any good. I write Python with NumPy, pandas and scikit-learn to compute evaluation metrics, and plot confusion matrices and ROC curves with Matplotlib and Seaborn.
Play with it on the right. Pick how good the model is and slide the decision threshold to see the trade-off between catching positives and raising false alarms. The Python tab shows the code behind it.
Machine Learning Foundations
Robotic Hand
A 3D-printed robotic hand that copies human finger movement. I engineered the electronics and the working mechanics.
It takes two kinds of input: a wearable glove with a flex sensor on every finger, or a laptop camera that tracks your fingers. An Arduino Uno turns each reading into an angle for one of five SG90 servos, and the servos pull tensioned fishing lines that run through the fingers like tendons.
- inputFlex-sensor glove, or camera hand tracking
- controlCustom Arduino Uno control system
- output5× SG90 servos on fishing-line tendons
Drag any finger down to bend it, or tap it. Each flex sensor has two leads into the breadboard, and one signal wire per finger runs to the Uno. Click a signal wire on the drawing to unplug it and that finger goes dead. The buttons under the glove bend one finger each.
On the real build a laptop camera tracks 21 points on your hand. Here you get the same view: drag a fingertip down to curl it.
input
arduino uno · analogRead
sg90 servos · angle
Project CAS
I co-developed the backend architecture and the desktop launcher applications for Project CAS, working across Node.js, Go and Tauri.
I'm also working in Unreal Engine to build features for CAS inside Fortnite.
Much of the job was making it run on real people's machines: diagnosing broken environments, execution-policy errors and systemic bugs until deployments were stable.
On the right is our actual launcher, rebuilt from the real Tauri app: the dev build, without the sign-in. Click around Home, the game pages, Friends, Settings and the Dev tools. Press ▢ to open it full screen.
- roleBackend co-developer
- builtBackend services, desktop launchers, Fortnite features in Unreal Engine
- fixedEnvironment issues, execution policy, systemic bugs
Fortnite artwork: portions of the materials used are trademarks and/or copyrighted works of Epic Games, Inc. All rights reserved by Epic. This material is not official and is not endorsed by Epic. This website is not endorsed by or affiliated with EA or its licensors.
Output
Languages
- Python
- Java
- Go
- JavaScript Node.js
- C / C++ Arduino
Frameworks & libraries
- scikit-learn
- pandas
- NumPy
- Matplotlib
- Seaborn
- Tauri
Hardware
- Arduino
- Microcontrollers
- Flex sensors
- Embedded troubleshooting
Systems & tools
- Linux
- Unreal Engine
- RAG learning