ML Engineer & Researcher

Ahmed Nasr

Senior Machine Learning Engineer working across computer vision, generative AI, and physics-informed machine learning. Based in Cairo, Egypt. Currently at Zazmic (Google Cloud Partner). M.Sc. Researcher at Cairo University.

Scientific ML Computational Intelligence Physics-Informed ML Computer Vision Deep Learning Generative AI & LLMs Signal Processing ML Systems
Get in Touch Explore My Work
Ahmed Nasr
Scroll

01 — Profile

About Me

MSc Researcher Physics-Informed ML Senior Machine Learning Engineer

I'm an ML Engineer and Researcher based in Cairo, Egypt, working across both industry and academia.

Currently a Senior Machine Learning Engineer at Zazmic (Google Cloud Partner), building Generative AI systems, LLM pipelines, agentic workflows, and computer vision solutions. Previously, I built production ML systems for the NEOM smart city project at Solutions by STC — deploying VLMs on multi-GPU clusters and architecting scalable ML serving infrastructure.

I also work as a freelance ML consultant on Upwork, taking on select projects in computer vision, speech/audio processing, NLP, and generative AI for international clients.

My M.Sc. research at Cairo University focuses on physics-informed ML models for solving partial differential equations — combining Gaussian Processes, VAEs, and Physics-Informed Neural Networks.

6+
Years of ML Engineering with international companies
M.Sc.
Computer Engineering
Cairo Univ. · GPA 3.6/4.0

Industries: Smart Cities · Healthcare · Generative AI · Autonomous Systems & Robotics

02 — Research & Focus Areas

Focus Areas & Research

I work across applied ML engineering and academic research. On the industry side, I build and deploy ML systems for production. On the research side, my M.Sc. focuses on physics-informed ML models for simulating PDEs and forward/inverse problems. I'm also interested in computer vision and signal processing.

◆ M.Sc. Thesis — Cairo University

Physics-Informed Generative Models for PDE Solving

Developing frameworks that combine Gaussian Process priors with Variational Autoencoders and physics-informed constraints for solving partial differential equations. Thesis in progress.

PINNs VAEs GPs Operator Learning Spectral Methods Fourier Features Kolmogorov-Arnold Networks (KAN)
∂

Physics-Informed ML

PINNs, operator learning, physics-constrained generative models, PDE solvers, spectral methods, and data-driven simulation of physical systems.

◎

Computer Vision

Object detection & tracking, segmentation, 3D LiDAR point clouds, VLMs, multi-modal fusion, edge deployment, and large-scale visual analytics.

∿

Signal Processing

Speech & audio recognition, image/video processing, biomedical signals, and audio classification for real-world applications.

⟁

Generative AI & LLMs

LLM pipelines, agentic orchestration (LangChain, LangGraph, Google ADK, MCP), RAG systems, structured output, and parameter-efficient fine-tuning.

⊕

ML Systems & Infrastructure

Production ML pipelines, multi-GPU serving (Ray Serve, vLLM), containerization, event-driven architectures, MLOps, and cloud/on-premise deployment.

◆ Roles I've held

Senior Machine Learning Engineer Senior Data Scientist ML Research Engineer Computer Vision Engineer ML Consultant

03 — Experience

Where I've Built Things

June 2025 — Present

Senior Machine Learning Engineer

Zazmic · United States (Remote) · Google Cloud Partner

  • Building and deploying Generative AI systems, including LLM pipelines, agentic workflows, and advanced computer vision solutions, with a focus on scalability, reliability, and real-world performance.
  • Architecting and delivering customer-facing AI solutions and POCs, translating business requirements into secure, production-ready systems deployed on Google Cloud infrastructure.
  • Partnering with project managers and cross-functional stakeholders to define technical strategy, align solutions with business objectives, and drive end-to-end implementation.
Dec 2023 — August 2025

Senior Machine Learning Engineer

Solutions by STC · Cairo, Egypt (Remote) · NEOM Smart City Project

  • Built production ML systems for NEOM across Vision, NLP, Speech, and Generative AI — serving intelligent analytics across millions of digital assets.
  • Deployed Vision-Language Models on multi-GPU clusters with Ray Serve & vLLM, with custom batching and autoscaling.
  • Architected scalable ML serving pipelines: async orchestration, event-driven inference, structured output enforcement, and containerized on-premise deployments.
  • Built CLIP Zero-shot classification, Whisper transcription, and image quality estimation systems.
April 2022 — Dec 2023

Computer Vision Engineer

COM-IOT Technologies · Dubai, UAE (Remote from Cairo)

  • Developed end-to-end ML systems for image and 3D LiDAR point cloud processing, powering smart city applications across security, traffic, and crowd analytics.
  • Built the company's ML/MLOps infrastructure from scratch — model development, versioning, evaluation, and deployment workflows for perception systems.
  • Optimized models for edge hardware deployment and shipped production-ready solutions across projects spanning several continents.
June 2021 — Present

Freelance Machine Learning Consultant

Upwork · Remote, Global Clients

  • Select consulting projects in Computer Vision, Audio/Speech, NLP, and Generative AI for international clients.
March 2021 — April 2022

Machine Learning R&D Engineer

Avelabs · Cairo, Egypt · Autonomous Vehicles

  • Built Deep Learning models for audio recognition on AutoHears — an AI-enabled audio sensor for autonomous vehicles.
  • Developed & optimized audio processing pipelines in Python/C++ and optimized models for hardware deployment using TensorFlow.
2019 — 2021

Research & Teaching — Ain Shams University

Multiple Roles · Cairo, Egypt

  • CV & ML Research Team Lead — Center of Mobility Research: Led research group developing computer vision-based traffic analysis tools.
  • AUV Perception Team Lead — ASMarine: Led team building the perception system for an Autonomous Underwater Vehicle (RoboSub 20, sponsored by Valeo). Detection, tracking, and ROS integration.
  • Teaching Assistant — Neural Networks course for senior Computer Engineering students.
  • Research Assistant — Autotronics Research Lab: Object detection, multi-object tracking, segmentation, and LiDAR-camera fusion on funded research projects.

04 — Skills

Technical Arsenal

ML & Deep Learning

Computer Vision Speech & Audio NLP Generative AI LLMs Transformers PINNs Operator Learning 3D Point Clouds Self-Supervised Learning LoRA Fine-tuning

Frameworks & Libraries

PyTorch HuggingFace TensorFlow Scikit-Learn Keras OpenCV Open3D PCL Librosa NumPy/SciPy Matplotlib Pandas

AI Agents & Orchestration

LangChain LangGraph Google ADK MCP Agentic Workflows RAG Semantic Search Vector DBs

Infrastructure & MLOps

Docker Ray Serve Ray vLLM Kafka FastAPI Flask MLFlow PyTorch Lightning HuggingFace Accelerate ONNX TensorRT OpenVINO LitServe

Cloud & Databases

GCP Vertex AI Cloud Run Cloud Functions BigQuery AWS (EC2, Lambda) MongoDB SQL Weaviate LanceDB Qdrant

Languages & Tools

Python C/C++ MATLAB Git Linux ROS LaTeX Jira

05 — Education

Academic Background

M.Sc. in Computer Engineering

Cairo University

Oct 2021 — Present  ·  GPA: 3.6 / 4.0

Thesis: Physics-Informed ML (PINNs, Operator Learning, Generative Models) for solving PDEs in forward and inverse problems and data-driven simulation. Coursework: Machine Intelligence II, Neural Networks, Cognitive Robotics, Computer Vision, Parallel Hardware & Computing, Advanced Security, Robotics II, Computer Architecture for Deep Learning, Advanced Deep Learning.

B.Sc. in Electrical & Computer Engineering

Ain Shams University

2015 — 2020

Major: Computer & Systems Engineering. Graduation Project: Developing the Perception System of Autonomous Underwater Vehicles (AUVs) for the RoboSub international competition — Project Grade: Excellent. Technically sponsored by Valeo.

06 — Publications

Research

A Rule Learning Approach for Building an Expert System to Detect Network Intrusions

Omar Galal, Ahmed Nasr, Lydia Wahid Rizkallah

International Journal of Intelligent Computing and Information Sciences, Vol. 23, Issue 1, 2023

07 — Training & Certificates

Summer Schools & Programs

2023

Oxford ML Summer School (OxML)

43 hours of lectures at the Oxford Mathematical Institute & Deep Medicine Program — ML fundamentals, advanced theory, and applications in health.

View Certificate →

2025

Intro to Computational Neuroscience

Summer school covering computational neuroscience foundations, fMRI brain decoding, and hands-on EEG analysis projects.

View Certificate →

08 — Beyond Work

Highlights

2020, 2021, 2022

HackZurich — 3× Participant

Participated three consecutive times in Europe's largest hackathon as an AI Developer — building AI prototypes with international teams in intensive coding marathons.

2017 — 2018

ROV MATE Competition

Ranked 4th in the MENA region for remotely operated underwater vehicles. Developed embedded software in C on ATmega and Python on Raspberry Pi.

09 — Contact

Let's Connect

Open to collaborations in ML, computer vision, signal processing, generative AI, physics-informed ML, and autonomous systems & robotics.

Email LinkedIn GitHub Phone Upwork