Skip to content
View Ahelsamahy's full-sized avatar

Block or report Ahelsamahy

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Ahelsamahy/README.md

Ahmed Mahfouz — Software Developer & AI Developer

AI Developer at Careweb and previously at VOIAR, with dual Computer Science and Telecom Engineering backgrounds. Passionate about building scalable AI-powered systems and meaningful digital tools—from healthcare-grade transcription pipelines to real-time robotics.

Creator of Storify - AI Story from an Image and ListiMate - Smart grocery list (1000+ cross-platform downloads). First-class graduate from Eszterházy Károly University (HU) and top-5 in Telecom Engineering at Delta University (EG).

🏢 Work Experience

💼 Careweb – Software Developer

Nov 2024 – Present · Utrecht - Netherlands 🇳🇱

  • Designed and implemented an AI-powered transcription and summarization pipeline using Whisper, FastAPI, vLLM, and Weaviate, reducing summarization time by 40% and enabling scalable 30k-token prompt handling.
  • Architected and deployed modular Java/Spring REST microservices as OSGi bundles on Apache Karaf.
  • Migrated legacy JSP interfaces to a modern bilingual (NL/EN) React-based front-end.
  • Built an observability stack using OpenTelemetry, Prometheus, Jaeger, and Grafana for GPU workload monitoring.

🤖 VOIAR (MWLC) – AI Developer

Mar 2023 – Nov 2024 · Utrecht - The Netherlands 🇳🇱

  • Developed edge-optimized YOLO vision models with GPU acceleration, cutting inference time from 980ms to 80ms.
  • Contributed to the official open-source Ultralytics YOLO repo by adding a TensorRT Docker image for Jetson Nano—boosting inference performance by 43%.
  • Implemented socket-based inter-robot communication protocols (ZMQ, TCP) with 0.0043s latency.
  • Built a mobile-first control interface deployed across multiple screen sizes.
  • Reduced redundant data in a 4M-record MySQL dataset by 75% per training cycle.

🧪 IoT Research Institute – Frontend Tester

Apr 2022 – Oct 2022 · Budapest - Hungary 🇭🇺

  • Conducted QA testing and validation for frontend modules in IoT-driven platforms.

🛠 Skills & Technologies

Languages
Python JavaScript TypeScript Java PHP

Frameworks & Libraries
NodeJS TailwindCSS jQuery NumPy

Tools & Infrastructure
Docker Nginx Postman Postgres

💻 My Posts

A minimal Docker image (JetPack 4.x) that uses CUDA 10.2, PyTorch 1.11, and TensorRT 8.2 to cut latency by ~43% and boost FPS by ~72%. It ended up being added to the main codebase of Ultralytics and is used by over 30K users.

How I turned "just an idea" into a smarter shopping tool

📫 Contact

📄 My full resume is available on my website.

LinkedIn Gmail

Pinned Loading

  1. MWLCDev/BYODR MWLCDev/BYODR Public

    Build Your Own Delivery Robot - Teleoperated, autonomous, lightweight and weatherproof. Free for personal use.

    Python 5 3

  2. Survive-Line Survive-Line Public

    Ball survives two sine waves using N.E.A.T algorithm

    TeX 2 1